Q&A

In this interview, Sujay Jadhav, global vice president, study start-up, Oracle Health Sciences, touches on how COVID has affected study start-up and what new perspectives it has forced the industry to have on its own challenges. [...]assessing site ability to leverage telehealth will be a factor in site selection. Andy Studna is an Assistant Editor for Applied Clinical Trials Sujay Jadhav Global Vice President, Study Start-Up, Oracle Health Sciences Problems with startup, more than any other phase of a clinical trial, have the greatest potential to increase timelines and budgets.

Patient Recruitment Goes High-Tech

Clinical trial patient recruitment is arguably the most difficult aspect of pharmaceutical development, because it involves a variety of factors beyond study sponsors' control. The aggregation of data across 80 hospitals and 20 systems, for the purpose of understanding patients, doing feasibility studies, or engaging in decentralized recruitment, is the trend we're seeing." Nimita Limaye, PhD, is the vice president of research for the life sciences R&D strategy and technology division at the International Data Corporation (IDC), a market research and advisory firm specializing in the technology industry and headquartered in Boston, Mass. Limaye says the rise of social media-based patient recruitment has opened the door for sponsors and investigators to mine real-world data and to give patients a more central focus in research.

Machine Learning and Deep Learning based Stock Market Prediction considering Covid-19 as a Feature

India's capital markets are witnessing intense uncertainty due to global market failures. Since the outbreak of COVID-19, risk asset prices have plummeted sharply. Risk assets declined half or more compared to the losses in 2008 and 2009. The high volatility is likely to continue in the short term;as a result, the Indian markets have declined sharply. In this paper, we have used different algorithms such as Gated Recurrent Unit, Long Short-Term Memory, Support Vector Regressor, Decision Tree, Random Forest, Lasso Regression, Ridge Regression, Bayesian Ridge Regression, Gradient Boost, and Stochastic Gradient Descent Algorithm to predict financial markets based on historical data available along with economic and financial features during this pandemic. According to our findings, deep learning models can accurately estimate financial indexes by utilizing non-linear transaction data. We found that the Gated Recurrent Unit performs better than the existing model. © 2023 IEEE.

Automating Biopharma Manufacturing

"With the advance of data science enabling factors such as easy access to scalable memory and computing resources;our growing competence in collecting, storing, and contextualizing data;advances in robotics;[and] the quickly evolving method landscape driven by the open-source community, the benefits of automation and simulation are becoming accessible in the notoriously complicated realm of biopharma manufacturing," says Marcel von der Haar, head of product strategy data analytics at Sartorius. "Plug-and-play" capabilities of automation systems, which enable flexible manufacturing and faster technology transfer, are more important than ever, he says. Walvax Biotech's new COVID-19 mRNA vaccine plant in China is another example of an intelligent and digital plant;it uses Honeywell's batch process control, building and energy management solution systems, and digital twins to monitor assets (5). "Automation brings in the data for machine learning to model the dynamic processes of cell growth and map it against the multiple dimensions provided by advanced sensors," explains Brandl.

Analysis and Prediction of Student Performance Based on Moodle Log Data using Machine Learning Techniques

During the COVID-19 Pandemic, many universities in Thailand were mostly locked down and classrooms were also transformed into a fully online format. It was challenging for teachers to manage online learning and especially to track student behavior since the teacher could not observe and notify students. To alleviate this problem, one solution that has become increasingly important is the prediction of student performance based on their log data. This study, therefore, aims to analyze student behavior data by applying Predictive Analytics through Moodle Log for approximately 54,803 events. Six Machine Learning Classifiers (Neural Network, Random Forest, Decision Tree, Logistic Regression, Linear Regression, and Support Vector Machine) were applied to predict student performance. Further, we attained a comparison of the effectiveness of early prediction for four stages at 25%, 50%, 75%, and 100% of the course. The prediction models could guide future studies, motivate self-preparation and reduce dropout rates. In the experiment, the model with 5-fold cross-validation was evaluated. Results indicated that the Decision Tree performed best at 81.10% upon course completion. Meanwhile, the SVM had the best result at 86.90% at the first stage, at 25% of the course, and Linear Regression performed with the best efficiency at the middle stages at 70.80%, and 80.20% respectively. The results could be applied to other courses and on a larger e-learning systems log that has similar student activity conditions and this could contribute to more accurate student performance prediction © 2023, International Journal of Emerging Technologies in Learning.All Rights Reserved.

The value of probabilistic forecasting in emergency medical resource planning under uncertainty

PurposeMost epidemic transmission forecasting methods can only provide deterministic outputs. This study aims to show that probabilistic forecasting, in contrast, is suitable for stochastic demand modeling and emergency medical resource planning under uncertainty.Design/methodology/approachTwo probabilistic forecasting methods, i.e. quantile regression convolutional neural network and kernel density estimation, are combined to provide the conditional quantiles and conditional densities of infected populations. The value of probabilistic forecasting in improving decision performances and controlling decision risks is investigated by an empirical study on the emergency medical resource planning for the COVID-19 pandemic.FindingsThe managerial implications obtained from the empirical results include (1) the optimization models using the conditional quantile or the point forecasting result obtain better results than those using the conditional density;(2) for sufficient resources, decision-makers' risk preferences can be incorporated to make tradeoffs between the possible surpluses and shortages of resources in the emergency medical resource planning at different quantile levels;and (3) for scarce resources, the differences in emergency medical resource planning at different quantile levels greatly decrease or disappear because of the existing of forecasting errors and supply quantity constraints.Originality/valueVery few studies concern probabilistic epidemic transmission forecasting methods, and this is the first attempt to incorporate deep learning methods into a two-phase framework for data-driven emergency medical resource planning under uncertainty. Moreover, the findings from the empirical results are valuable to select a suitable forecasting method and design an efficient emergency medical resource plan.

Predicting customer churn based on changes in their behavior patterns

Customer retention is one of the most important tasks of a business, and it is extremely important to allocate retention resources according to the potential profitability of the customer. Most often the problem of predicting customer churn is solved based on the RFM (Recency, Frequency, Monetary) model. This paper proposes a way to extend the RFM model with estimates of the probability of changes in customer behavior. Based on an analysis of data relating to 33 918 clients of a large Russian retailer for 2019-2020, it is shown that there are recurring patterns of change in their behavior over a single year. Information about these patterns is used to calculate the necessary probability estimates. Incorporating these data into a predictive model based on logistic regression increases prediction accuracy by more than 10% on the metrics AUC and geometric mean. It is also shown that this approach has limitations related to the disruption of behavioral patterns by external shocks, such as the lockdown due to the COVID-19 pandemic in April 2020. The paper also proposes a way to identify these shocks, making it possible to forecast degradation in the predictive ability of the model.

Conjoint Analysis based Predictive Analytics to study the employee attitude towards virtual and digital work practices in IoT work environments

Covid-19 has been one of the most disruptive pandemics to date. Among the other aspects of disruption, it also disrupted the way people work in organizations. Many of the organizations surrendered their offices for good. However, there are many ill effects of these unconventional work practices also. This research study aims to explore the perception of the employees towards the adoption of Virtual and flexible work practices. The study uses a conjoint analysis approach on different possible Work Practice Profiles, that specify the nature of work (Virtual, offline, or hybrid), nature of work schedule (flexible, or fixed), nature of ownership (individual, or team), and length of working hours (8.5 hours, or 9.5 hours or 10.5 hours). The study finds that the number of working hours is the most important criterion for the employees followed by mode of work, responsibility, and work schedule. © 2023 IEEE.

Capturing complexity over space and time via deep learning: An application to real-time delay prediction in railways

Predictive analytics is an increasingly popular tool for enhancing decision-making processes but is in many business settings based on rule-based models. These rule-based models reach their limits in complex settings. This study compares the performance of a rule-based system with a customised LSTM encoder-decoder deep learning model for predicting train delays. For this, we use a purposefully built real-world dataset on railway transportation, where trains' interdependence over the network makes delay prediction more difficult. Results show that the deep learning model, which incorporates rich spatiotemporal interdependency information in real-time, outperforms the rule-based system by 18%, with the difference increasing to above 23% with higher complexity. The study also dissects the performance difference across different settings: dense versus rural areas, peak versus off-peak hours, low versus high delay, and before versus during the COVID-19 pandemic. The deep learning model is implemented as a proof of concept for decision support within Belgium's railway infrastructure company Infrabel. © 2023 Elsevier B.V.

Predictive Analytics of Stock Market as a Time Series

This paper presents a new methodology and a comparative study using past stock market data that can help businesses take investing or divesting decisions in critical situations in the future. These may be like the COVID-19 pandemic, where market volatility is extremely high, thus creating an urgent need for better decision support systems to minimise loss and ensure better profits. The results of the study are based on the comparison of different configurations of ARIMAX, Prophet, LSTM and Bidirectional LSTM Models trained on historical NSE data. By understanding the correlation and variations in the data processing and model training parameters, we have successfully proposed a LSTM neural network model training and optimising method which could successfully help businesses take both long and short term profitable decisions before and after big financial and market crises with a respective accuracy of 98.60 percent and 96.97 percent. © 2022 IEEE.

Smart Meter Data Analytics for Building Monitoring System: A Case Study

Decarbonization, decentralization, and digitalization are the prominent paths for the energy sector in the future. The rise of smart meters across consumers, and industries led to a massive collection of fine-grained energy and electricity consumption-related data. A data science challenge is to analyze the Smart Meter data for the benefit of both the energy providers and the consumers. In this paper, An attempt has been made to analyze the smart meter collected from the IIT Hyderabad campus and presented the analysis into descriptive, predictive, and prescriptive analytics. The data collected from more than 50 meters over a period of one year have been analyzed and results obtained. Interesting trends such as the impact of COVID-19 on campus energy consumption have been examined. The framework for energy data analytics presented in this paper will be useful for any campus in general, and the recommendations presented will save energy expenses. © 2023 IEEE.

Machine Learning Techniques for Predicting Risks of Late Delivery

Supply chain is a cornerstone of the eCommerce industry and is a key component in its growth. Supply chain data analytics and risk management in the eCommerce space have picked up steam in recent times. With the availability of suitable & capable resources for big data and artificial intelligence, predictive analytics has become a significant area of interest to achieve organizational excellence by exploiting data available and developing data-driven support systems. The existing literature in supply chain risk management explain various methods assisting to identify & mitigate risks using big data and machine learning (ML) techniques across industries. Although ML techniques are used in various industries, not many aspects of eCommerce had utilized predictive analytics to their benefit. In the eCommerce industry, delivery is paramount for the business. During COVID-19 pandemic, needs changed. Reliable delivery services are preferred to speedy delivery. Multiple parameters involve delivering the product to a customer as per promised due date. This research will try to predict the risks of late deliveries to online shopping customers by analyzing the historical data using machine learning techniques and comparing them by multiple performance metrics. As a part of this comparative study, a new hybrid technique which is a combination of Logistic Regression, XGBoost, Light GBM, and Random Forest is built which has outperformed all the other ensemble and individual algorithms with respect to accuracy, specificity, precision, and F1-score. This study will benefit the eCommerce companies to improve their customer satisfaction by predicting late deliveries accurately and early. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

Machine Learning Based Predictive Analytics for Agriculture Inventory Management System

When the globe was hit by the vicious Covid 19 pandemic, multiple industries faced the virus's wrath and that included the agricultural warehouse industry. Consequently, many warehouses which had received large shipment stocks of agricultural products were never to be used again as it had reached its expiration date. This led to major losses for the agricultural warehouses as well as losses in crops for farmers and large scale agriculturists. The main objective of this paper is to build a model which utilises 3 heavy-weight algorithms (Seasonal Autoregressive Integrated Moving Average-SARIMA, Long short term memory-LSTM and Holt Winters) and predicts the agricultural needs of retailers and consumers based on previous data from different warehouses. Deploying this system will not help in the regulation of goods in warehouses but will also aid in maximizing the profits and minimizing the losses for warehouses. The algorithm with the least MAE(Mean Absolute Error) value will be considered for forecasting the sales of the aforementioned product. © 2022 IEEE.

Empirical Analysis of Psychological Well-Being of Students During the Pandemic with Rebooted Remote Learning Mode

The direct and indirect mental health stressors, especially associated with the "tele-burden of pandemic” added due to the adoption of the remote learning paradigm, have led to increased online fatigue, distress, and burnout. This research aims to comprehend the perception of psychological distress experienced by Indian students placed in the new online learning setting. Subsequently, the observed symptomatology is used to predict the student's susceptibility toward developing specific psychological challenges. Primarily, a phenomenological study is conducted on 732 student participants to understand their psychological well-being during this ongoing COVID-19 crisis. Subsequently, machine learning is used to train a model with learned features from the data extracted to detect six psychological states, amusement, neutral, low stress, high stress, depression, and anxiety. Two supervised machine learning algorithms, namely random forest and artificial neural network, are used to perform the predictive analytics of psychological well-being. Experimental evaluation reports a classification accuracy of 90.4% for the random forest and 89.15% for the neural network. The qualitative research findings help foster the need to look for coping strategies involving counselors and psychologists to decrease the risk of psychological distress and preserve students' psychological health and well-being in the current setting. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

AI in Tourism: Leveraging Machine Learning in Predicting Tourist Arrivals in Philippines using Artificial Neural Network

Tourism is one of the most prominent and rapidly expanding sectors that contribute significantly to the growth of a country's economy. However, the tourism industry has been most adversely affected during the coronavirus pandemic. Thus, a reliable and accurate time series prediction of tourist arrivals is necessary in making decisions and strategies to develop the competitiveness and economic growth of the tourism industry. In this sense, this research aims to examine the predictive capability of artificial neural networks model, a popular machine learning technique, using the actual tourism statistics of the Philippines from 2008-2022. The model was trained using three distinct data compositions and was evaluated utilizing different time series evaluation metrics, to identify the factors affecting the model performance and determine its accuracy in predicting arrivals. The findings revealed that the ANN model is reliable in predicting tourist arrivals, with an R-squared value and MAPE of 0.926 and 13.9%, respectively. Furthermore, it was determined that adding training sets that contain the unexpected phenomenon, like COVID-19 pandemic, increased the prediction model's accuracy and learning process. As the technique proves it prediction accuracy, it would be a useful tool for the government, tourism stakeholders, and investors among others, to enhance strategic and investment decisions © 2023, International Journal of Advanced Computer Science and Applications.All Rights Reserved.

Prediction Model of hospitalization time of COVID-19 patients based on Gradient Boosted Regression Trees.

When an outbreak of COVID-19 occurs, it will cause a shortage of medical resources and the surge of demand for hospital beds. Predicting the length of stay (LOS) of COVID-19 patients is helpful to the overall coordination of hospital management and improves the utilization rate of medical resources. The purpose of this paper is to predict LOS for patients with COVID-19, so as to provide hospital management with auxiliary decision-making of medical resource scheduling. We collected the data of 166 COVID-19 patients in a hospital in Xinjiang from July 19, 2020, to August 26, 2020, and carried out a retrospective study. The results showed that the median LOS was 17.0 days, and the average of LOS was 18.06 days. Demographic data and clinical indicators were included as predictive variables to construct a model for predicting the LOS using gradient boosted regression trees (GBRT). The MSE, MAE and MAPE of the model are 23.84, 4.12 and 0.76 respectively. The importance of all the variables involved in the prediction of the model was analyzed, and the clinical indexes creatine kinase-MB (CK-MB), C-reactive protein (CRP), creatine kinase (CK), white blood cell count (WBC) and the age of patients had a higher contribution to the LOS. We found our GBRT model can accurately predict the LOS of COVID-19 patients, which will provide good assistant decision-making for medical management.

Observation of Online vs Offline Learning Experience

As the outbreak of COVID-19 increased in various countries. India is also majorly affected with the COVID-19 by that education system is affected, and it has transferred the traditional face-to-face teaching to online education platform. Considering student's perspective on both online and offline learning mode in India, we conducted a survey to collect the data. In that survey questionnaire, focus was on the factors and situation which can affect the education system. Using that data, we used Kruskal Wallis test to collect the evidence for which learning mode is better and Naive Bayes Algorithm, we were able to conclude the results. © 2022 IEEE.

Systematic Review of Qualitative and Quantitative Studies on Perceived Employability of Graduates

The assessment of future students' employability by the Institute of Higher Learning in collaboration with career centres is one of the most crucial steps in the educational industry for establishing an active and ascendable plan. Predictive analysis for this project is done using machine learning. This study investigates the Employability Signals of Undergraduates in accordance with the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) criteria. The findings demonstrate that higher education was where the most accurate predictor of undergraduate students' employability was initially examined. The study's conclusions can be used to develop a roadmap that will make it simpler to use predictive analytics. The findings of this study may also facilitate the creation and application of predictive analytics, one of the possible approaches for analysing the education data gathered during the pre-covid period for this study. Systematic literature reviews should be trustworthy, repeatable, and valid when used in scientific investigations. As a result, the inquiry will reach a conclusion based on the evaluations found on pertinent and customized dates. © 2023 IEEE.

AI: new source of competitiveness in higher education

Purpose>This paper aims to survey the current landscape of artificial intelligence (AI) applications in higher education institutions (HEIs) and recommend future directions.Design/methodology/approach>This paper reviews the recent trends, showcases the applications and provides future directions through a review of current uses of AI in HEIs.Findings>The results of this study highlight successful applications of AI technologies in three main areas of college operation: student learning experience;student support;and enrollment management.Research limitations/implications>This review has important implications for early adopters of AI by HEIs in providing a competitive advantage. The limitation lies in the scope of the review. It is not comprehensive and does not cover other areas of college operations.Originality/value>This is the first review about AI in higher education. It is of value in building future research and serving as a framework for AI applications in HEI.

Forecasting Overtime Budgets for Naval Fleet Maintenance Facilities Using Time-Series Analysis During Transient System States

In 2019, predictive models were initially developed to attempt to better predict an annual budget for staffing overtime hours within a Royal Canadian Navy (RCN) fleet maintenance facility. The H20.ai open-source framework was used, and models were implemented in the R programming language. Model validation at the time showed the predicted hours were within 5% error rate compared to the actual data. However, when it came to re-apply the process to fiscal year 2020/2021 data, the impact of the COVID-19 pandemic on factors such as the workforce and the logistics supply chain, changed the system dynamics sufficiently that the autoML algorithms had difficulty generating accurate estimates. Therefore, it was decided to examine how times series forecasting methods would predict overtime hours at the fleet maintenance facility. Since historical daily data were readily available, the open-source Prophet model developed by Facebook was used because it can incorporate multiple seasonal patterns, as well as variable holiday effects. The models were tested on fiscal years 2019/2020 and 2020/2021, which showed over 90% accuracy in predicting the total overtime hours. The revised approach in this follow-on study was used to provide financial comptrollers with a prediction for fiscal year 2021/2022. © 2023, Springer Nature Switzerland AG.