Milos Jovanovic | University of Belgrade (original) (raw)

Uploads

Papers by Milos Jovanovic

Research paper thumbnail of Gaussian Conditional Random Fields for Classification

arXiv (Cornell University), Jan 31, 2019

Research paper thumbnail of Generic

Research paper thumbnail of A Framework for Integrating Domain Knowledge in Logistic Regression with Application to Hospital Readmission Prediction

International Journal on Artificial Intelligence Tools, 2019

It is commonly understood that machine learning algorithms discover and extract knowledge based o... more It is commonly understood that machine learning algorithms discover and extract knowledge based on data at hand. However, a huge amount of knowledge is available which is in machine-readable format and ready for inclusion in machine learning algorithms and models. In this paper, we propose a framework that integrates domain knowledge in form of ontologies/hierarchies into logistic regression using stacked generalization. Namely, relations from ontology/hierarchy are used in stacking manner in order to obtain higher, more abstract concepts. Obtained concepts are further used for prediction. The problem we solved is unplanned 30-days hospital readmission, which is considered as one of the major problems in healthcare. Proposed framework yields better results compared to Ridge, Lasso, and Tree Lasso Logistic Regression. Results suggest that the proposed framework improves AUC by up to 9.5% on pediatric datasets and up to 4% on morbidly obese patients’ datasets and also improves AUPRC b...

Research paper thumbnail of Applying interpolative Boolean algebras in medical sciences – a case study

Research paper thumbnail of FAIR: Fair Adversarial Instance Re-weighting

arXiv (Cornell University), Nov 15, 2020

Research paper thumbnail of An Investigation of Human Trajectories in Ski Resorts

Analyzing human trajectories based on sensor data is a challenging research topic. It has been an... more Analyzing human trajectories based on sensor data is a challenging research topic. It has been analyzed from many aspects like clustering, process mining, and others. Still, less attention has been paid on analyzing this data based on hidden factors that drive the behavior of people. We, therefore, adapt the standard matrix factorization approach and reveal factors which are interpretable and soundly explain the behavior of a dynamic population. We analyze the motion of a skier population based on data from RFID-recorded ski entrances of skiers on ski lift gates. The approach is applicable to other similar settings, like shopping malls or road traffic. We further applied recommender systems algorithms for testing how well we can predict the distribution of ski lift usage (number of ski lift visits) based on hidden factors, but also on other benchmark algorithms. The matrix factorization algorithm showed to be the best recommender score predictor with an RMSE of 2.569 ± 0.049 and an ...

Research paper thumbnail of Fair classification via Monte Carlo policy gradient method

Engineering Applications of Artificial Intelligence, 2021

Research paper thumbnail of A white box approach in modeling phase of the data mining process

Research paper thumbnail of WhiBo-RapidMiner plug-in for component based data mining algorithm design

ABSTRACT WhiBo is a framework for defining, running and testing of machine-learning algorithms as... more ABSTRACT WhiBo is a framework for defining, running and testing of machine-learning algorithms as a composition of reusable components (i.e. white-box design approach). These components are extracted from well known algorithms as well as their partial improvements for solving specific sub-problems. WhiBo is intended for data mining practitioners, algorithm designers and machine-learning software developers. Currently the WhiBo framework consists of a reusable components (RC) repository, a RC-based generic decision tree algorithm, GUI for design of RC-based algorithms and an operator group for performance and significance testing of decision trees. Additionally, a generic partitioning clustering algorithm is included in WhiBo.

Research paper thumbnail of Internal Evaluation Measures as Proxies for External Indices in Clustering Gene Expression Data

Research paper thumbnail of A method for design of data-tailored partitioning algorithms for optimizing the number of clusters in microarray analysis

Research paper thumbnail of Using data mining on student behavior and cognitive style data for improving e-learning systems: a case study

International Journal of Computational Intelligence Systems, 2012

Research paper thumbnail of White-Box or Black-Box Decision Tree Algorithms: Which to Use in Education?

IEEE Transactions on Education, 2013

ABSTRACT University students are usually taught data mining through black-box data mining algorit... more ABSTRACT University students are usually taught data mining through black-box data mining algorithms, which hide the algorithm's details from the user and optionally allow parameter adjustment. This minimizes the effort required to use these algorithms. On the other hand, white-box algorithms reveal the algorithm's structure, allowing users to assemble algorithms from algorithm building blocks. This paper provides a comparison between students' acceptance of both black-box and white-box decision tree algorithms. For these purposes, the technology acceptance model is used. The model is extended with perceived understanding and the influence it has on acceptance of decision tree algorithms. An experiment was conducted with 118 senior management students who were divided into two groups-one working with black-box, and the other with white-box algorithms-and their cognitive styles were analyzed. The results of how cognitive styles affect the perceived understanding of students when using decision tree algorithms with different levels of algorithm transparency are reported here.

Research paper thumbnail of An architecture for component-based design of representative-based clustering algorithms

Data & Knowledge Engineering, 2012

ABSTRACT We propose an architecture for the design of representative-based clustering algorithms ... more ABSTRACT We propose an architecture for the design of representative-based clustering algorithms based on reusable components. These components were derived from K-means-like algorithms and their extensions. With the suggested clustering design architecture, it is possible to reconstruct popular algorithms, but also to build new algorithms by exchanging components from original algorithms and their improvements. In this way, the design of a myriad of representative-based clustering algorithms and their fair comparison and evaluation are possible. In addition to the architecture, we show the usefulness of the proposed approach by providing experimental evaluation.

Research paper thumbnail of Reusable components in decision tree induction algorithms

Computational Statistics, 2011

Research paper thumbnail of Reusable components for partitioning clustering algorithms

Artificial Intelligence Review, 2009

Clustering algorithms are well-established and widely used for solving data-mining tasks. Every c... more Clustering algorithms are well-established and widely used for solving data-mining tasks. Every clustering algorithm is composed of several solutions for specific sub-problems in the clustering process. These solutions are linked together in a clustering algorithm, and they define the process and the structure of the algorithm. Frequently, many of these solutions occur in more than one clustering algorithm. Mostly, new

Research paper thumbnail of Business intelligence system development over document meta data in the organization

… -časopis za teoriju i …, 2010

Poslovna inteligencija je aktivno područje istraživanja i primene. Iako je oblast mlada, nudi dos... more Poslovna inteligencija je aktivno područje istraživanja i primene. Iako je oblast mlada, nudi dosta rešenja, prvenstveno namenjena boljem izveštavanju i analizi koji će podržati proces donošenja odluke, na svim nivoima odlučivanja. Kako je proces odlučivanja prisutan u svakoj oblasti ...

Research paper thumbnail of Component-based decision trees for classification

Intelligent Data Analysis, 2011

Research paper thumbnail of Modeling of slowly changing dimensions of a data mart for monitoring parameters in the educational process

Research paper thumbnail of Evolutionary approach for automated component-based decision tree algorithm design

Intelligent Data Analysis

Research paper thumbnail of Gaussian Conditional Random Fields for Classification

arXiv (Cornell University), Jan 31, 2019

Research paper thumbnail of Generic

Research paper thumbnail of A Framework for Integrating Domain Knowledge in Logistic Regression with Application to Hospital Readmission Prediction

International Journal on Artificial Intelligence Tools, 2019

It is commonly understood that machine learning algorithms discover and extract knowledge based o... more It is commonly understood that machine learning algorithms discover and extract knowledge based on data at hand. However, a huge amount of knowledge is available which is in machine-readable format and ready for inclusion in machine learning algorithms and models. In this paper, we propose a framework that integrates domain knowledge in form of ontologies/hierarchies into logistic regression using stacked generalization. Namely, relations from ontology/hierarchy are used in stacking manner in order to obtain higher, more abstract concepts. Obtained concepts are further used for prediction. The problem we solved is unplanned 30-days hospital readmission, which is considered as one of the major problems in healthcare. Proposed framework yields better results compared to Ridge, Lasso, and Tree Lasso Logistic Regression. Results suggest that the proposed framework improves AUC by up to 9.5% on pediatric datasets and up to 4% on morbidly obese patients’ datasets and also improves AUPRC b...

Research paper thumbnail of Applying interpolative Boolean algebras in medical sciences – a case study

Research paper thumbnail of FAIR: Fair Adversarial Instance Re-weighting

arXiv (Cornell University), Nov 15, 2020

Research paper thumbnail of An Investigation of Human Trajectories in Ski Resorts

Analyzing human trajectories based on sensor data is a challenging research topic. It has been an... more Analyzing human trajectories based on sensor data is a challenging research topic. It has been analyzed from many aspects like clustering, process mining, and others. Still, less attention has been paid on analyzing this data based on hidden factors that drive the behavior of people. We, therefore, adapt the standard matrix factorization approach and reveal factors which are interpretable and soundly explain the behavior of a dynamic population. We analyze the motion of a skier population based on data from RFID-recorded ski entrances of skiers on ski lift gates. The approach is applicable to other similar settings, like shopping malls or road traffic. We further applied recommender systems algorithms for testing how well we can predict the distribution of ski lift usage (number of ski lift visits) based on hidden factors, but also on other benchmark algorithms. The matrix factorization algorithm showed to be the best recommender score predictor with an RMSE of 2.569 ± 0.049 and an ...

Research paper thumbnail of Fair classification via Monte Carlo policy gradient method

Engineering Applications of Artificial Intelligence, 2021

Research paper thumbnail of A white box approach in modeling phase of the data mining process

Research paper thumbnail of WhiBo-RapidMiner plug-in for component based data mining algorithm design

ABSTRACT WhiBo is a framework for defining, running and testing of machine-learning algorithms as... more ABSTRACT WhiBo is a framework for defining, running and testing of machine-learning algorithms as a composition of reusable components (i.e. white-box design approach). These components are extracted from well known algorithms as well as their partial improvements for solving specific sub-problems. WhiBo is intended for data mining practitioners, algorithm designers and machine-learning software developers. Currently the WhiBo framework consists of a reusable components (RC) repository, a RC-based generic decision tree algorithm, GUI for design of RC-based algorithms and an operator group for performance and significance testing of decision trees. Additionally, a generic partitioning clustering algorithm is included in WhiBo.

Research paper thumbnail of Internal Evaluation Measures as Proxies for External Indices in Clustering Gene Expression Data

Research paper thumbnail of A method for design of data-tailored partitioning algorithms for optimizing the number of clusters in microarray analysis

Research paper thumbnail of Using data mining on student behavior and cognitive style data for improving e-learning systems: a case study

International Journal of Computational Intelligence Systems, 2012

Research paper thumbnail of White-Box or Black-Box Decision Tree Algorithms: Which to Use in Education?

IEEE Transactions on Education, 2013

ABSTRACT University students are usually taught data mining through black-box data mining algorit... more ABSTRACT University students are usually taught data mining through black-box data mining algorithms, which hide the algorithm's details from the user and optionally allow parameter adjustment. This minimizes the effort required to use these algorithms. On the other hand, white-box algorithms reveal the algorithm's structure, allowing users to assemble algorithms from algorithm building blocks. This paper provides a comparison between students' acceptance of both black-box and white-box decision tree algorithms. For these purposes, the technology acceptance model is used. The model is extended with perceived understanding and the influence it has on acceptance of decision tree algorithms. An experiment was conducted with 118 senior management students who were divided into two groups-one working with black-box, and the other with white-box algorithms-and their cognitive styles were analyzed. The results of how cognitive styles affect the perceived understanding of students when using decision tree algorithms with different levels of algorithm transparency are reported here.

Research paper thumbnail of An architecture for component-based design of representative-based clustering algorithms

Data & Knowledge Engineering, 2012

ABSTRACT We propose an architecture for the design of representative-based clustering algorithms ... more ABSTRACT We propose an architecture for the design of representative-based clustering algorithms based on reusable components. These components were derived from K-means-like algorithms and their extensions. With the suggested clustering design architecture, it is possible to reconstruct popular algorithms, but also to build new algorithms by exchanging components from original algorithms and their improvements. In this way, the design of a myriad of representative-based clustering algorithms and their fair comparison and evaluation are possible. In addition to the architecture, we show the usefulness of the proposed approach by providing experimental evaluation.

Research paper thumbnail of Reusable components in decision tree induction algorithms

Computational Statistics, 2011

Research paper thumbnail of Reusable components for partitioning clustering algorithms

Artificial Intelligence Review, 2009

Clustering algorithms are well-established and widely used for solving data-mining tasks. Every c... more Clustering algorithms are well-established and widely used for solving data-mining tasks. Every clustering algorithm is composed of several solutions for specific sub-problems in the clustering process. These solutions are linked together in a clustering algorithm, and they define the process and the structure of the algorithm. Frequently, many of these solutions occur in more than one clustering algorithm. Mostly, new

Research paper thumbnail of Business intelligence system development over document meta data in the organization

… -časopis za teoriju i …, 2010

Poslovna inteligencija je aktivno područje istraživanja i primene. Iako je oblast mlada, nudi dos... more Poslovna inteligencija je aktivno područje istraživanja i primene. Iako je oblast mlada, nudi dosta rešenja, prvenstveno namenjena boljem izveštavanju i analizi koji će podržati proces donošenja odluke, na svim nivoima odlučivanja. Kako je proces odlučivanja prisutan u svakoj oblasti ...

Research paper thumbnail of Component-based decision trees for classification

Intelligent Data Analysis, 2011

Research paper thumbnail of Modeling of slowly changing dimensions of a data mart for monitoring parameters in the educational process

Research paper thumbnail of Evolutionary approach for automated component-based decision tree algorithm design

Intelligent Data Analysis