Improving hospital readmission prediction using individualized utility analysis

Dear Editor, The costs of antipsychotic drugs (APDs) used in the treatment of mental disorders with psychosis are mentioned in treatment guidelines (APA 2021, NICE 2014). While the American Psychiatric Association guideline states that every specialist should make decisions according to the rules and conditions of their country and their region, the National Institute of Health and Clinical Excellence guideline emphasizes that drug costs must be taken into consideration in the treatment process. Classical or first-generation antipsychotic drugs (FAPDs) are relatively cheaper in terms of sales prices compared to atypical or second-generation antipsychotic drugs (SAPDs) with a slightly different effect mechanism. The price difference between the two drug groups can be so large that sometimes it may be necessary to consider whether the cost of a second-generation drug is worth its benefit. While deciding on the use of first-generation or second-generation drugs, a multifaceted assessment should be made, such as the patient's level of compliance with the treatment, the possibility of occurrence of side effects, the possible effects of these side effects on body health and treatment compliance, and whether or not the costs are covered. The most important criterion that determines the choice of medication for psychiatrists is of course the multi-dimensional benefit/harm ratio that the drug used will reveal in the long term. We think that in our country, which, in terms of economic indicators is not in a strong position as an importer of pharmaceutical raw materials from abroad, APDs' cost calculation should be considered because drug costs constitute an important part of the direct treatment costs of psychotic disorders in developing countries such as Turkey (Yıldız and Cerit 2006). We calculated the unit (mg) price based on the box prices of the APDs in use in 2020, thinking that it might work when calculating the cost of the illness using APDs as the main component of the treatment and calculated the annual average drug costs with the daily average dosage. Although the daily treatment dose varies with the stage of the illness and the individual characteristics of the patient, the average doses recommended for maintenance treatment were used here (Öztürk and Ulusahin 2018). The daily and annual cost calculations based on the assumption that the average maintenance treatment dose was used with the unit price obtained from (Drug Prices 2020) the drugs in the Turkish pharmaceutical market in September 2020 are shown in Table 1. A similar study was done in 2005 (Yıldız 2005). The purpose of this article is to redetermine the average costs of APDs in the Turkish pharmaceutical market every 15 years and to bring them to the attention of experts in terms of cost-effectiveness studies. When the costs in 2005 are examined, it is seen that the annual costs of the FAPDs were around 450 TRY, and the annual cost of oral preparations of SAPDs was 2,500 TRY (5 times the first generation). In 2005, there was only one depot of SAPD (risperidon consta) that allowed intramuscular (IM) administration, and its average annual cost was 5,400 TRY, 3 times more than the tablet form (1,700 TRY). In 2005, when the price of risperidone consta, which was the first second-generation depot APD, were compared with the prices of the first-generation depot drugs (fluphenazine = 380 TRY, flupentixol = 876 TRY, zuclopentixol = 730 TRY), the cost difference was 6-14 times. This almost-10-fold difference between the cost of the first and second generation APDs was remarkable. It is seen that this difference (risperidone consta = 10,807 TRY, fluphenazine = 916 TRY, flupentixol = 1,007 TRY, zuclopenthixol = 2,372 TRY, and haloperidol deconate 237 TRY) did not change in 2020. In 2020, the average RETHINKING THE COST OF ANTIPSYCHOTIC TREATMENT: THE AVERAGE COST OF THE DRUGS USED IN TURKEY IN 2020 2 Türk Psikiyatri Dergisi 2 Turkish Journal of Psychiatry Letter to the Editor 146 147 annual cost of oral use preparations of FAPDs is 925 TRY, while the average annual cost of oral forms of SAPDs is 2,580 TRY. The 5-fold difference observed in 2005 between the first and second-generation ones of the oral APDs decreased to 2.5 times in 2020. It is clear that while the difference between the cost of oral use of first- and second-generation drugs was halved in 2020, the difference between the costs of depot preparations applied with IM did not change. In 2005, the average dollar rate was 1.34 TRY, and in 2020 it was 7.02 TRY (Republic of Turkey Central Bank Exchange Rates, 2021). It is understood that the 5-fold increase in dollar exchange rate is not reflected in all drug prices in the same way. For example, there was a 3 to 4-fold increase in the prices of haloperidol, chlorpromazine, fluphenazine, trifluperazine and zuclopenthixol, while a less than two-fold increase in pimozide, flupenthixol, sulpiride, amisulpride and quetiapine and a decrease in the prices of clozapine, olanzapine, ziprasidone and risperidone in the tablet form. There is also a two-fold increase in the price of risperidone consta. The fluctuations in drug prices in 2005 and 2020 are shown in Table 2 in 500, 1,000, 2,000, 3,000 and 5,000 TRY brackets. It is noteworthy that while some drugs have moved into an upper price bracket in terms of annual costs, some have fallen into a lower price bracket. The prices of the second generation long-acting (depot) antipsycotic drugs (LA-APDs), which were not available in the Turkish pharmaceutical market in 2005, are quite high compared to others. In 2020, the annual cost of all of them, including risperidone consta, is over 10 thousand TRY. It is understood that the underlying reason for such price increase is the fact that the drug is wanted/sought after/new/marketed rather than the dollar exchange rate. For example, while there was a certain increase in the price of FAPDs, the increase in the price of some of the SAPDs (sulpiride, amisulpride, quetiapine tablet) was low, while the price of some others (clozapine, olanzapine, ziprasidone, risperidone tablet) decreased. It should also be taken into account that the effect of generic drugs entering the market during this period may have had an impact on price changes. It is noteworthy that while the annual cost of risperidone consta was approximately 3 times higher than the tablet form (5,400 TRY versus 1,700 TRY) in 2005, this difference reached 14 folds (10,807 TRY versus 742 TRY) in 2020. In 2005, the difference between the lowest daily cost (0.07 TRY) and the highest daily cost (14.80 TRY) was 211 times (Yıldız 2005), this difference had receded to 111 times (0.35 TRY versus 38.72 TRY) in 2020. Still a huge difference, isn't it? Table 1. Current Forms, Box Prices, Daily and Annual Costs in For Maintenance Treatment of Antipsychotic Drugs Available in the Pharmaceutical Market in September 2020 in Turkey No Generic name Trade name Dosage forms (mg) BV Price# TRY/Mg ADD Cost/d Cost/y 2005** 1 Haloperidol Norodol 5, 10, 20 tb 5/50 17.57 0.070 5 0.35 127 26 5, 10 amp 5/5 5.35 0.214 5 1.07 390 - 50, 150 LAI 50/1 9.80 0.196 1/15* 0.65 237 - 2 Chlorpromazine Largactil 25,100 tb 100/30 17.92 0.006 300 1.79 653 197 3 Fluphenazine Prolixin 25 LAI 25/1 17.57 0.703 1/7* 2.51 916 380 4 Trifluoperazine Stilizan 1, 2, 5 drj; 1 amp 5/30 14.52 0.096 10 0.97 354 91 5 Pimozide Nörofren 2 tb 2/30 19.33 0.322 4 1.29 470 365 6 Flupenthixol Fluanxol 3 drj 3/50 65.75 0.438 6 2.63 960 526 20 LAI 20/1 19.33 0.966 1/7* 2.76 1,007 876 7 Zuklopenthixol Clopixol 2, 10, 25 tb 2/50 38.65 0.386 20 7.72 2,817 701 200 LAI, 50 acu 200/1 45.55 0.227 1/7* 6.50 2,372 730 8 Sulpirid Dogmatil 200 tb 200/24 23.15 0.005 600 3.00 1,095 876 9 Amisulpirid Solian 200 tb 200/60 146.92 0.012 600 7.20 2,628 2,387 10 Quetiapine Seroquel 25, 50, 100, 200, 300, 400 tb 300/30 137.17 0.015 600 9.00 3,285 2,628 11 Clozapine Leponex 25, 100 tb 100/50 32.56 0.006 400 2.40 876 1,898 12 Olanzapine Zyprexa 5, 10, 20 tb 10/28 152.96 0.546 10 5.46 1,992 2,606 13 Ziprasidone Zeldox 20, 40, 60, 80 tb 60/56 189.89 0.056 120 6.72 2,452 3,541 14 Sertindole Serdolect 4, 12, 16, 20 tb 16/28 453.53 1.012 16 16.19 5,909 - 15 Risperidone Risperdal 1, 2, 3, 4 tb; 1 sol 2/20 20.34 0.508 4 2.03 741 1,719 Ris. Consta 25, 37.5, 50 LAI 37.5/1 444.17 11.840 1/15* 29.61 10,807 5,402 16 Paliperidone Invega 3, 6, 9 tb 6/28 213.15 1.268 6 7.61 2,777 - Xeplion 50, 75, 100, 150 LAI 100/1 1161.56 11.615 1/30* 38.72 14,132 - Trevicta 175, 263, 350, 525 LAI 350/1 3426.95 9.788 1/90* 38.08 13,899 - 17 Aripiprazole Abilify 5, 10, 15, 20 tb; 1 sol 20/28 113.25 0.404 20 8.08 2,949 - Abilify Main. 400 LAI 400/1 971.17 2.420 1/30* 32.37 11,815 - BV: Baseline value (in mg of the form and the number in the box), Price#: Box price of the base value in TRY, TRY/mg: Value per milligram in Turkish Lira, ADD: Average daily dose, Cost/d: Daily cost in TRY, Cost/y: Annual cost in TRY, mg: Milligram, tb: Tablet, drj: Dragee, amp: Ampoule, LAI: Long-acting injectable, acu: Acuphase, d: Day, TRY: Turkish Lira, *LAI per 7,15,30 or 90 days, **Annual cost in TRY in 2005. 148 Received: 14.01.2021, Accepted: 31.03.2021, Available Online Date: 07.01.2022 1Prof., 2Res. Assis., Kocaeli University School of Medicine, Department of Psychiatry, Kocaeli, Turkey. e-mail: myildiz60@yahoo.com https://doi.org/10.5080/u26315 The difference in 2005 between oral FAPDs prices and SAPDs prices seems to have halved in 2020. In 2020, the average daily treatment cost of oral drugs, whether for the first generation or the second generation, is 3 TRY (approximately the same for FAPDs applied with IM), while the daily cost of LA-SAPDs is around 33 TRY. It is seen that the difference between costs is approximately 11 times. This difference increases to 50 times for haloperidol deconate. From here, the following judgment can be made: in order for LA-SAPDs to be preferred, they must be at a value that will constitute at least 11 times

BackgroundAtrial fibrillation (AF) ablation is an effective treatment for reducing episodes and improving quality of life in patients with AF. However, long-term AF-free rates after AF ablation are inconsistent across the population, ranging from 50% to 75%. Patient selection relies on individual clinical assessment, highlighting a critical gap in population-level predictive analytics. While existing risk scores (eg, CHADS₂ [congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, and stroke], CHA₂DS₂-VASc [congestive heart failure, hypertension, age ≥75 years, diabetes mellitus, stroke, vascular disease, age, and sex category], CAAP-AF [coronary artery disease, left atrial diameter, age, AF, antiarrhythmic drugs, and female sex category]) have been applied to predict AF ablation outcomes, their performance in administrative claims data remains unclear. Leveraging large administrative claims databases represents an opportunity to develop standardized, scalable prediction models that could inform population health management and resource allocation at a national level.ObjectiveThis study utilizes machine learning (ML) models on claims data to explore if integrating International Classification of Diseases (ICD) billing codes outperforms traditional stroke and AF risk scores in predicting 1-year AF ablation outcomes.MethodsWe analyzed claims data from the Merative MarketScan Research Medicare database (2013‐2020) to identify 14,521 patients who underwent AF ablation. To predict 1-year AF-free outcomes, we developed logistic regression and extreme gradient boosting (XGBoost) models using demographic characteristics, comorbidity indices, and ICD diagnostic codes from the 2 years preceding ablation. Model predictions were compared with claims-based implementations of established risk scores—CHADS2, CHA2DS2-VASc, and a modified CAAP-AF (without left atrial diameter and the number of failed antiarrhythmic drugs). The ML models were also assessed on subgroups of patients with paroxysmal AF, persistent AF, and both AF and atrial flutter from October 2015 onward.ResultsAmong 14,521 patients (mean age 71.5, SD 5.31 y; n=5800, 39.94% female), AF ablation success occurred in 54.01% (n=7843). XGBoost achieved areas under the receiver operating characteristic curve (AUCs) of 0.528, 0.521, and 0.529 for the whole, female, and male AF ablation groups, respectively, and better discrimination than CHADS2, CHA2DS2-VASc, and the modified CAAP-AF in all AF ablation groups (whole population, female, and male). While CHA2DS2-VASc and the modified CAAP-AF showed higher recall (>0.798), their precision (<0.540) was lower than XGBoost (0.552‐0.556). In subgroup analyses of International Classification of Disease, Tenth Revision (ICD-10) patients (n=7646), the models incorporating ICD codes demonstrated better performance than those using only demographic and comorbidity data across most AF subtypes, with the highest AUC (0.544) observed in patients with paroxysmal AF.ConclusionsWhile the ML models achieved statistically significant improvements over claim-based implementations of established clinical risk scores (AUC 0.528‐0.544 vs 0.498‐0.505), the modest predictive performance highlights challenges in predicting procedural outcomes using administrative data that lack key clinical variables (eg, left atrial size and medication details). Our findings establish that while standardized outcome prediction using nationally available administrative data is technically feasible, current performance is insufficient for clinical decision-making and better suited for health system quality monitoring and comparative effectiveness research applications.

Machine learning models (MLMs) used to influence surgical decision making often require large and complex datasets upon which to train. However, there is a paucity of literature pertaining to the ability to apply standard MLMs to small ICU datasets within resource-constrained institutions. ML models were applied to a prospective cohort of critically ill mechanically ventilated patients from a teaching hospital in rural Kenya. Characteristics included an ICU scoring system specifically for resource-constrained environments (Tropical Intensive Care Score (TropICS)). Outputs included AUC of the ROC and the feature importance table. Python-based MLMs included XGBoost and KNN. AUC of the ROC was calculated to predict mortality as the primary endpoint. There were 294 patients, with an average age of 40.2years, 64.3% male, 23.8% trauma, and an overall mortality of 60.2%. With respect to mortality patients who died were older (43.5 versus 35years; p < 0.001), but with no difference in male gender (64.8% versus 63.8%; p = 0.9), or having been transferred from outside facilities (34% versus 21.5%; p = 0.5). Whilst there was no difference in the rate of tachycardia or acidosis, patients who died were more likely to present with hemodynamic instability (31% versus 6%; p < 0.001) and higher clinical severity scores. In predicting mortality, the ML models performed very well (XGBoost AUC = 0.82). Within MLM feature importance, the Tropical Intensive Care Score (TropICS) performed as well as APACHE-II and the SAPS. ML models can be effectively applied to a small ICU dataset within resource-constrained environments. ML models must demonstrate functionality prior to incorporating within prospective clinical predictive models. Contextualized ICU scoring systems (TropICS) performed well within MLMs.

Background: The air quality of any area depends upon the various PMs (particulate matter) and hazardous gases present in the air. Low-cost PM sensors and gas sensors are present in different target places to monitor the air quality, read the environmental data, and transmit it to local servers through the IoT device. The low-cost sensor is not reliable due to its low sensing capacity; therefore, the read data is calibrated with the meteorological data presented by the nearby meteorological Centre of that particular area. The calibrated reading data sent to the server could be analyzed through some Machine Learning [ML] models. The ML models help to predict the risk of asthma in a particular area. The risk of asthma is directly related to the air quality of the surroundings. It is observed that the air quality of the industrial area is much worse than the non-industrial belt. Air quality monitoring of industrial areas is always a challenging task due to the ununiformed pollution in some segregated places around the industry, emitting pollutants mostly from chimneys. The air quality of any area depends upon the PM (PM), i.e., PM2.5 and PM10.0, as well as the gasses like NO2(Nitrogen Dioxide), NH3 (Ammonia), SO2(Sulfur dioxide), CO(Carbon monoxide), O3(Ozone) and Benzene. These are the most hazardous gases generally emitted by common heavy industries like iron and steel. In this article, the researchers considered the industrial belt of the Asansol- Durgapur region of West Bengal, India, and predicted the risk of asthma attacks for the test dataset. The experiment was carried out on 10 different supervised machine learning [SML] models as well as semi-supervised machine learning (SSML) models. The SML models have been further refined through hyper-parameter tuning, and better results have been obtained in the case of some ML models. The result has been compared with the existing literature considering the same external environment from where the meteorological data was collected, and similar ML models have been used. The research outperformed the existing literature, which is depicted in the result and analysis section of the article. Methods: The study evaluated ML models, both supervised and semi-supervised, to assess pollution levels. Relevant features were selected while less relevant ones were discarded. Accuracy levels of different ML algorithms werecompared in the results. The most effective model for an IoT system was chosen to maximize accuracy. In semi-supervised learning, feature selection followed supervised learning, but testing was akin to unsupervised learning. Results were compared with supervised learning data, enhancing reliability. Results: The result employing various classifiers werepresented across tables after the independent parameter Ozone was removed. Following the output of several classifiers, the results were verified using the k-fold validation method, with k being set to 5 or 10, accordingly. Tables display the best outcome, which is indicated in bold characters. method: In this research work the researcher considered 9 different ML models and used them as supervised as well as semi supervised model to determine the pollution level of the certain area. In this research work the researcher also selected the most relevant features and discarded the less relevant features. In case of SML algorithm, the accuracy level of the different ML algorithm has been determined and depicted in the result analysis section. The most effective ML model has been chosen for the proposed embedded system so that accuracy could be achieved at most. In case of semi supervised algorithm the feature selection is done as per the supervised algorithm. In this case the training is done same as the SML algorithm, but the testing phase is done like unsupervised machine learning algorithm where the decision parameter is predicted and ultimately matched with the previously achieved data of SML algorithm. The reliability of this approach is much more effective than simple SML algorithm. Conclusion: This study focused on predicting asthma risk in the Asansol-Durgapur industrial belt, India, using low-cost PM and gas sensors. Data calibration with meteorological inputs enhanced accuracy. ML models predicted risk and were refined through hyper-parameter tuning. Comparative analysis showed superior performance, emphasizing the importance of precise air quality monitoring. While offering a robust framework for future research, the study’s limitation lies in its area-specific dataset.

Critical infrastructure protection and national security are enhanced by the security and reliability of networks. Various types of information circulate on these networks, ranging in classification from open to closed. The consequences of cyberattacks on these networks can be severe, including reputational damage, financial loss, operational disruption and data leakage. Traditional security methods, such as firewalls and anti-virus software, are becoming less effective against modern and ever-changing cyber threats. As a result, powerful network intrusion detection systems (IDS) have become indispensable for proactive detection and mitigation of cyber attacks. Machine learning has become a viable method for creating adaptive intrusion detection tools that can detect new and complex attack patterns. By learning from huge labelled network traffic datasets, ML models can understand the subtle patterns and differentiating features of normal and abnormal or malicious traffic flows. This allows them to detect possible cyber threats and intrusions that traditional signature-based IDSs cannot detect. Extracting discriminative features and training appropriate classification models from such data is a challenging task. In the presented study, we analyse the effectiveness of ML algorithms for detecting cyberattacks, in particular distributed denial of service (DDoS) attacks, in network traffic data. In the presented study, a network attack detection system is developed using ML and deep learning (DL) models and experimented on the CICIDS2017 dataset. The main objectives of the study are to develop a strategy for extracting valuable information from raw network streams; to study the impact of data preparation on the false positive rate; and to conduct a comparative analysis of ML models for cyberattack detection. The main goal of the study is to provide an understanding of the development of a reliable adaptive network intrusion detection system using ML approaches that increase cybersecurity capabilities and protect against future cyberattacks.

The computed tomography-severity score (CT-SS) quantifies the severity of pulmonary involvement and is significantly associated with disease severity, intensive care unit (ICU) admissions, and mortality in coronavirus disease-2019 (COVID-19) patients. There is very limited information on the prognostic value of CT-SS when used in machine learning (ML) models to predict ICU admission in COVID-19 patients. In this study, the prognostic significance of CT-SS in ML model-based prediction of ICU admission among COVID-19 patients was evaluated. In this retrospective study, a hospital-based database from 6,854 COVID-19 patients was reviewed. To evaluate the prognostic significance of CT-SS in predicting ICU admission in patients, seven ML methods were trained separately using the most important features, with and without CT-SS data, and their performances were compared. After applying exclusion criteria, 815 COVID-19 patients remained. Just over half of the patients (54.85%) were male, and the mean age was 57.22±16.76 years. The CT-SS index was the strongest predictor among the parameters examined, and integrating this index into the training dataset enhanced ML model performance. The k-nearest neighbors model with 93.3% accuracy, 97.3% sensitivity, 89.4% specificity, and an area under the curve of approximately 98.8% showed the best performance for predicting ICU admission in COVID-19 patients. The results showed that CT-SS is a key predictor for ML models of ICU admission in COVID-19 patients. The ML models developed using a dataset including CT-SS are efficient risk stratification tools for identifying critical COVID-19 patients, thereby facilitating optimal allocation of limited hospital resources.

As dependency on credit cards and online transactions is increasing, the scope for unauthorized payments or fraudulent transactions is on the rise. Huge financial losses are borne by banks every year due to fraudulent transactions. The process of identifying such transactions among millions of transaction records can be perfected using machine learning models, even more so by ensembling them. This study takes a real-time financial dataset and analyzes the efficiency of various machine learning models in predicting fraudulent transactions. Individual ML models are also analysed, ensembled and tested on the dataset. Upon extensive analysis and research, it is concluded that an ensembled model of LightGBM, decision tree classifier and XGBoost gave the highest AuC score of 87%. Further, a user-interactive tool is also created, which predicts fraudulent transactions from real-time transactions. The tool makes use of the model with highest prediction efficiency and presents it in an interactive manner. This study and practical demonstration show advancement in prediction of fraudulent transactions when compared to existing ML models and its user-friendly software incorporation.KeywordsFraud detectionEnsembleAccuracyFraudulent transactionBankingMachine learning model

Customer attrition is especially an issue in industries such as retail, banking, and telecommunications where customer acquisition costs are significantly higher than the costs of retaining repeat customers. The customer lack of interest is now predictable through machine learning models, and deep learning has become instrumental in early intervention for retention. In order to assess the quality of churn prediction, the study tests six basic machine learning techniques: random forest, logistic regression, and the k-nearest neighbors method, as well as four deep learning techniques: long short term memory (LSTM), bidirectional LSTM, convolutional neural networks (CNN), and artificial neural networks (ANN). The performance of the model is then assessed via the evaluation matrices, including the accuracy, precision, recall, and F1-score from the customer's behavioral data after feature extraction from large datasets. The study reveals that DL models offer improved handling of the churn and non-churn customer classification and Random Forest as well as other ML models comparable accuracy. This research can conclude that LSTM and ANN models outshine in actual-world churn prediction circumstances, especially when long-term consumer behavior evaluation is required. To enhance the current outcomes of a given prediction model, this research focuses on data preprocessing and the utilization of bootstrapping, feature extraction, and the combination of multiple models. The implications of the study provide specific practical recommendations for firms to effectively manage customer churn and increase customer retention by employing data-dealing techniques.

Machine learning models are often black box and thus practitioners find it difficult to predict their behavior, debug, and implement specific improvements. The paper is a proposal of a no-code comprehension platform, Model Insights, offering detailed analysis and insights with actionable recommendations on machine learning and deep learning models, without the need of code knowledge on part of the end-user. The platform automatically processes uploaded models in order to comprehend their architecture, measure performance metrics, offer explainability via SHAP-driven visualizations and generate individualized improvements advice. Model Insights is written in a modern technology stack, React.js frontend with TypeScript and Tailwind CSS, and Python backends with TensorFlow/Keras and scikit-learn, and can serve multiple model formats (.h5, .pkl, .joblib) and data formats (e.g. tabular and image data). The system uses a variety of explainability measures, such as SHAP-based feature importance scores, individual prediction explanations, and overall performance measures, such as ROC curves, confusion matrices, and precisions-recall. Moreover, the site has an integrated chatbot assistant when submitting ML/DL queries, and version comparison to monitor performance change in a model. User testing shows that the platform helps greatly to improve model transparency and interpretability and allows both a beginner and expert user to recognize areas of poor performance, confirm or exclude model decisions, and make informed corrections. The findings indicate that model understanding and development efficacy is significantly improved, which makes Model Insights a viable solution in responsible AI creation and model management in both an academic and enterprise setting.

Resource allocation is playing a vital role in grid environment because of the dynamic and heterogeneous nature of grid resources. Literature offers numerous studies and techniques to solve the grid resource allocation problem. Some of the drawbacks occur during grid resource allocation are low utilization, less economic reliability and increased waiting time of the jobs. These problems were occurred because of the inconsiderable level in the code of allocating right resources to right jobs, poor economic model and lack of provision to minimize the waiting time of jobs to get their resources. So, all these drawbacks need to be solved in any upcoming resource allocation technique. Hence in this paper, the efficiency of the resource allocation mechanism is improved by proposing two allocation models. Both the allocation models have used the Genetic Algorithm to overcome all the aforesaid drawbacks. However, one of the allocation models includes penalty function and the other does not consider the economic reliability. Both the models are implemented and experimented with different number of jobs and resources. The proposed models are compared with the conventional resource allocation models in terms of utilization, cost factor, failure rate and make span.

Background Measles remains a major public health issue, especially in regions with low vaccination rates and limited healthcare access. Timely outbreak prediction is crucial for effective intervention and resource allocation. Traditional epidemiological models struggle with real-time predictions due to reliance on historical data and limited analysis. Artificial intelligence (AI) offers a data-driven approach to capturing complex outbreak patterns and improving prediction accuracy. Methods This study developed and evaluated machine learning (ML) models to predict measles outbreaks at the community level, using real-world data collected by community volunteers in Ethiopia. The dataset includes demographic, vaccination status, location, clinical, and image data, providing a comprehensive assessment of outbreak risk factors. Five ML models-Random Forest (RF), Support Vector Machine (SVM), Logistic Regression (LR), Decision Tree (DT), and K-Nearest Neighbors (KNN)-were assessed using metrics such as area under the receiver operating curve (AUC), precision, recall, and F1-score. Results The results demonstrate that machine learning models can effectively predict measles outbreaks, with the top-performing model, K-Nearest Neighbors (KNN), achieving an AUC of 0.87 (95% CI: 0.84-0.90). Among the five models tested, KNN outperformed the others, showing the highest predictive accuracy, and the difference in AUC was statistically significant (p &amp;lt; 0.05). Conclusions The model's ability to capture complex epidemiological factors underscores its potential to enhance surveillance and early warning systems. This research highlights the effectiveness of ML-based measles prediction in resource-limited settings. Integrating these models into public health strategies can improve early detection and reduce measles impact. Future work will focus on refining the model, incorporating real-time data, and expanding evaluation metrics for practical deployment. Key messages • Machine learning enables accurate early measles detection fusing different data sources, supporting faster response and better health outcomes in communities. • Machine learning offers a powerful tool for predicting outbreaks early, filling critical gaps in low-resource health systems.

Background: Coronary Artery Disease (CAD) is one of the biggest causes of mortality worldwide. Risk stratification for early detection is essential for the primary prevention of CAD. QRISK3 is known to overestimate future CAD risk in some populations, resulting in unnecessary preventive treatment and reduced cost-effectiveness and safety. Combining machine learning model with the metaheuristic optimisation approach using the PSO algorithm may outperform QRISK3 in predicting CAD. It improves performance by selecting the best-performing subset of features related to clinical outcomes. Research Question: Does the performance of Machine Learning Models Combined with PSO Algorithm for Feature Selection as a Metaheuristic Optimisation Approach in Predicting Coronary Arterial Disease using the UK Biobank dataset outperform the QRISK3 calculator? Aims: This study is to assess the accuracy of QRISK3 in predicting CAD using the UK Biobank dataset. It aims to evaluate the efficacy of machine learning models on the identical dataset for predicting CAD. The work utilises the PSO algorithm for feature selection to identify the optimal subset of features from the UK Biobank dataset. Methods: This study utilises data from the UK Biobank. The dataset consists of 348,015 participants aged 24-84 with no prior diagnosis of CAD. The performance of both QRISK3 and machine learning models was evaluated separately using ROC analysis. Several machine learning models were employed: Logistic Regression, Decision Tree, Random Forest, Naïve Bayes, and Gradient Boosting. The dataset was split into training and test sets with a ratio of 4:1 for the machine learning models. Each model has been developed by adding a PSO algorithm to enhance the model's classification accuracy. Results: Out of the total 348,015 participants, 23,136 individuals (6.64%) were diagnosed with CAD within 10 years following their first visit, while 324,879 individuals (93.4%) did not develop CAD. The AUC value of the QRISK3 prediction is 0.6113, while the combined machine learning models using the PSO algorithm using the gradient boosting model achieve an AUC of 0.7258, showing better performance. Conclusions: This study shows hybrid machine learning models optimised with the PSO algorithm can better predict CAD than QRISK3. These ML models can effectively identify high-risk CAD patients, allowing for more personalised preventative strategies and supporting policymakers in implementing lifestyle change recommendations.

Interpretable vs. noninterpretable machine learning models for data-driven hydro-climatological process modeling

Our goal was to assess the postoperative 90-day hospital costs of patients with prolonged air leak (PAL) including costs incurred after discharge from the initial index hospitalization. We performed a retrospective analysis of 982 patients undergoing lobectomy (898) or segmentectomy (78) (April 2014-August 2018). A total of 167 operations were open, 780 were video-assisted thoracoscopic surgery and 28 were robotic. A PAL was defined as an air leak >5 days. The 90-day postoperative costs included all fixed and variable costs incurred during the 90 days following surgery. The postoperative costs of patients with and without PAL were compared. The independent association of PAL with postoperative 90-day costs was tested after adjustment for patient-related factors and other complications by a multivariable regression analysis. PAL occurred in 261 patients (27%). Their postoperative stay was 4 days longer than that of those without PAL (9.6 vs 5.7; P < 0.0001). Compared to patients without PAL, those with PAL had 27% higher index postoperative costs [7354€, standard deviation (SD) 7646 vs 5759€, SD 7183, P < 0.0001] and 40% higher 90-day postoperative costs (18 340€, SD 23 312 vs 13 102€, SD 10 264; P < 0.0001). The relative postoperative costs (the difference between 90-day and index postoperative costs) were 50% higher in PAL patients compared to non-PAL patients (P < 0.0001) and accounted for 60% of the total 90-day costs. Multivariable regression analysis showed that PAL remained an independent factor associated with 90-day costs (P < 0.0001) along with the occurrence of other cardiopulmonary complications (P < 0.0001), male gender (P = 0.018), low carbon monoxide lung diffusion capacity (P = 0.043) and thoracotomy approach (P = 0.022). PAL is associated not only with increased index hospitalization costs but also with increased costs after discharge. Evaluation of the cost-effectiveness of measures to prevent air leaks should also include post-discharge costs.

This paper studies the performance of classical and quantum machine learning models for sentiment analysis task. Here, popular machine learning algorithms viz support vector machine (SVM), gradient boosting (GB) and random forest (RF) are compared with variational quantum classifier (VQC) using two sets of parameterized circuits viz EfficientSU2 and RealAmplitudes. For experimenting with VQC, IBM Quantum Experience and IBM Qiskit were used while for classical machine learning models, scikit-learn was used. It was found that the performance of the VQC was slightly better than popular machine learning algorithms. For our experiments, we have used popular restaurant sentiment analysis dataset. The extracted features from this dataset and then after applying PCA reduced the feature set into 5 features. Quantum ML models were trained using 100 epochs and 150 epochs. Overall, four Quantum ML models were trained and three Classical ML models were trained. The performance of the trained models was evaluated using standard evaluation measures viz, Accuracy, Precision, Recall, F-Score etc. In all the cases EfficientSU2 based model with 100 Epochs performed better than all other models. Efficient SU2 model with 100 epochs produced an accuracy of 74.5% and an F-Score of 0.7605 which were highest across all the trained models.