A5020050774- Imbalanced Data Classification Techniques
- Financial Distress and Bankruptcy Prediction
- Adversarial Robustness in Machine Learning
- Bayesian Modeling and Causal Inference
- Fuzzy Systems and Optimization
- Advanced Data Processing Techniques
- Advanced Statistical Methods and Models
- Advanced Computational Techniques in Science and Engineering
- Explainable Artificial Intelligence (XAI)
- Credit Risk and Financial Regulations
- Machine Learning in Healthcare
- Multi-Criteria Decision Making
- Neural Networks and Applications
- Machine Learning and Data Classification
- Statistical Methods and Inference
Warsaw University of Technology
2021-2023
Machine learning has proved to generate useful predictive models that can and should support decision makers in many areas. The availability of tools for AutoML makes it possible quickly create an effective but complex model. However, the complexity such is often a major obstacle applications, especially terms high-stake decisions. We are experiencing growing number examples where use black boxes leads decisions harmful, unfair or simply wrong. In this paper, we show very simplify without...
Abstract We focus on modelling categorical features and improving predictive power of neural networks with mixed numerical in supervised learning tasks. The goal this paper is to challenge the current dominant approach actuarial data science a new architecture network training algorithm. key proposal use joint embedding for all features, instead separate entity embeddings, determine representation which fed, together other into hidden layers target response. In addition, we postulate that...
Rapid development of advanced modelling techniques gives an opportunity to develop tools that are more and accurate. However as usually, everything comes with a price in this case, the pay is loose interpretability model while gaining on its accuracy precision. For managers control effectively manage credit risk for regulators be convinced quality too high. In paper, we show how take scoring analytics next level, namely present comparison various predictive models (logistic regression,...
Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI
The majority of automated machine learning (AutoML) solutions are developed in Python, however a large percentage data scientists associated with the R language. Unfortunately, there limited available. Moreover high entry level means they not accessible to everyone, due required knowledge about (ML). To fill this gap, we present forester package, which offers ease use regardless user's proficiency area learning. is an open-source AutoML package implemented designed for training high-quality...