Individual Machine Learning Models Research Articles

Genetic Algorithm Based Optimized Feature Engineering and Hybrid Machine Learning for Effective Energy Consumption Prediction

Smart grids are developing rapidly, leading to the need for accurate forecasts of power consumption. However, developing a precise time series model for energy forecasting is difficult. It has to be trained using optimal meteorological features such as temperature and time lags to qualify for a beneficial model. We have proposed an approach that uses an ensemble machine learning model based on XGBoost, support vector regressor (SVR), and K-nearest neighbors (KNN) regressor algorithms. We have also used the genetic algorithm (GA) to predict total load consumption from optimal feature selection. Using Jeju island's electricity consumption data as a case study shows that the proposed ensemble model optimized with GA is more accurate than the individual machine learning models. Using only the best-selected weather and time features, the proposed model records all the features of a complicated time series and shows a reduction in the mean absolute percentage error (MAPE) and the root mean square log error for the week ahead forecasts. We got 3.35 % MAPE of the three months test data by applying the proposed model. The smart grids operators can manage resources effectively to provide excellent services to the consumers based on the recommended model outcomes.

A living environment prediction model using ensemble machine learning techniques based on quality of life index

The living environment is an area that has the features necessary for all people to keep living their lives. Nevertheless, as time has progressed, needs and economic capacities of people have changed; consequently, the living environment has begun to express different meanings for different people. This differentiation brings the following question to mind: Can every person live in the same living environment under the same conditions? Of course not! People differ from each other in their preferences, needs, economic capacities, and many other aspects, so they must be assigned living environments that best fit their unique features. In this study, machine learning (ML) techniques and ensemble machine learning (EML) techniques are used for the establishment of a living environment prediction model. The data required for the quality of life index are obtained from a questionnaire. This questionnaire was prepared to consider people’s desires and economic capacities. Quality of life index is computed taking the weighted arithmetic mean of questionnaire data. Calculated quality of life indexes of the individuals is assigned to available quality of life classes of the cities. ML-based classification techniques have been used for prediction of assigned classes of the cities. A living environment prediction model is proposed in this study by using an ML and EML techniques-based quality of life index. For predicting living environment, ML-based methodologies include an artificial neural network (ANN), support vector machines (SVM) and EML-based methodologies include stacking and voting. The prediction results of ensemble models are better than prediction results of individual ML models, especially the stacking-based model composed of SMO + LMT and which reached the best performance values using the 80% split method.

Machine-Learning Based Stacked Ensemble Model for Accurate Analysis of Molecular Dynamics Simulations.

Accurate, faster, and on-the-fly analysis of the molecular dynamics (MD) simulations trajectory becomes very critical during the discovery of new materials or while developing force-field parameters due to automated nature of these processes. Here to overcome the drawbacks of algorithm based analysis approaches, we have developed and utilized an approach that integrates machine-learning (ML) based stacked ensemble model (SEM) with MD simulations, for the first time. As a proof-of-concept, two SEMs were developed to analyze two dynamical properties of a water droplet, its contact angle, and hydrogen bonds. The two SEMs consisted of two layered networks of random forest, artificial neural network, support vector regression, Kernel ridge regression, and k-nearest neighbors ML models. The root-mean-square error values, uncertainty quantification, and sensitivity analysis of both the SEMs suggested that the final result was more accurate as compared to that of the individual ML models. This new computational framework is very general, robust, and has a huge potential in analyzing large size MD simulation trajectories as it can capture critical information very accurately.

Enhanced Predictive Models for Construction Costs: A Case Study of Turkish Mass Housing Sector

The analysis of a construction project, regarding cost, is one of the most vital problem in planning. Due to its nature, the construction sector is an area of strong competition and estimation works are of vital importance. In recent years the Turkish Republic has started a serious urban regeneration movement in parallel to its economic development. This study is based on the drawings and quantities of 63 detached multi-story reinforced concrete housing unit projects of the Housing Development Administration (TOKI) and the Turkey Residential Building Cooperative Union (TURKKONUT). TOKI is a public company and its projects are that have been applied to 282 separate projects and are being applied to a further 266. On the other side TURKKONUT is a union of 1347 private building cooperative and have been completed 200,000 residential building. The main objective of this study is to improve the estimation accuracy of individual machine learning techniques, namely multi-layer perceptron and classification and regression trees and compares the performance of two machine learning meta-algorithms (i.e., bagging and random subspace) on a real world construction cost estimation problem. The study shows that the estimation accuracy of ensemble models are better than the models that constructed by their base learners and ensemble models may improve individual machine learning models.

Mapping groundwater contamination risk of multiple aquifers using multi-model ensemble of machine learning algorithms

Constructing accurate and reliable groundwater risk maps provide scientifically prudent and strategic measures for the protection and management of groundwater. The objectives of this paper are to design and validate machine learning based-risk maps using ensemble-based modelling with an integrative approach. We employ the extreme learning machines (ELM), multivariate regression splines (MARS), M5 Tree and support vector regression (SVR) applied in multiple aquifer systems (e.g. unconfined, semi-confined and confined) in the Marand plain, North West Iran, to encapsulate the merits of individual learning algorithms in a final committee-based ANN model. The DRASTIC Vulnerability Index (VI) ranged from 56.7 to 128.1, categorized with no risk, low and moderate vulnerability thresholds. The correlation coefficient (r) and Willmott's Index (d) between NO3 concentrations and VI were 0.64 and 0.314, respectively. To introduce improvements in the original DRASTIC method, the vulnerability indices were adjusted by NO3 concentrations, termed as the groundwater contamination risk (GCR). Seven DRASTIC parameters utilized as the model inputs and GCR values utilized as the outputs of individual machine learning models were served in the fully optimized committee-based ANN-predictive model. The correlation indicators demonstrated that the ELM and SVR models outperformed the MARS and M5 Tree models, by virtue of a larger d and r value. Subsequently, the r and d metrics for the ANN-committee based multi-model in the testing phase were 0.8889 and 0.7913, respectively; revealing the superiority of the integrated (or ensemble) machine learning models when compared with the original DRASTIC approach. The newly designed multi-model ensemble-based approach can be considered as a pragmatic step for mapping groundwater contamination risks of multiple aquifer systems with multi-model techniques, yielding the high accuracy of the ANN committee-based model.

Ensemble Learning Approach for Probabilistic Forecasting of Solar Power Generation

Probabilistic forecasting accounts for the uncertainty in prediction that arises from inaccurate input data due to measurement errors, as well as the inherent inaccuracy of a prediction model. Because of the variable nature of renewable power generation depending on weather conditions, probabilistic forecasting is well suited to it. For a grid-tied solar farm, it is increasingly important to forecast the solar power generation several hours ahead. In this study, we propose three different methods for ensemble probabilistic forecasting, derived from seven individual machine learning models, to generate 24-h ahead solar power forecasts. We have shown that while all of the individual machine learning models are more accurate than the traditional benchmark models, like autoregressive integrated moving average (ARIMA), the ensemble models offer even more accurate results than any individual machine learning model alone does. Furthermore, it is observed that running separate models on the data belonging to the same hour of the day vastly improves the accuracy of the results. Getting more accurate forecasts will help the stakeholders come up with better decisions in resource planning and control when large-scale solar farms are integrated into the power grid.

Bagging ensemble models for bank profitability: An emprical research on Turkish development and investment banks

The purpose of this study is to find the determinants of the profits for the Development and Investment Banks (IaDB) in Turkey. In Turkish Banking System, the main financial source of the banks is the deposits, which constitute almost%60 of the balance sheet. As being a sub-group of the banking system, IaDB are not allowed to accept deposits in Turkey, which changes the total structure of the profitability compared to other banks. Till today, none of the relevant research was concentrated on the profit structure of the IaDB neither in Turkey nor in any other countries. Such research would fill that unexpectedly disregarded yet highly important gap.Therefore, to address this gap, quarterly financial data (10 balance sheet ratios) of 13 banks in the period of 2002Q4-2014Q3 were utilized. As a profit measurement among all other available measures, Return on Equity was chosen as dependent variable as it was the most used one as well as many other researcher have preferred as well. This study investigates the potential usage of bagging (Bag), which is one of the most popular ensemble learning methods, in building ensemble models, is used to predict the determinants of Turkish IaDB profitability. Three well-known tree-based machine learning (ML) models (i.e., Decision Stump (DStump), Random Tree (RTree), Reduced Error Pruning Tree (REPTree)) are deployed as base learner. This empirical study indicates that bagging ensemble models (i.e., Bag-DStump, Bag-RTree, Bag-MLP and Bag-REPTree) are superior to their base learners and could improve the prediction accuracy of individual ML models (i.e., DStump, RTree, REPTree).

A novel ensemble scheme for the multi-focus image fusion using support vector machine

In the multi-focus image fusion problem, the source images are obtained from the same scene. They are fused to get an image that contains all well-focussed objects. Previously, individual machine-learning models are proposed for image fusion. The performance of individual models is limited to fuse the useful information extracted from the blurred images. To address this problem, we developed a novel ensemble scheme for multi-focus image fusion using support vector machines (SVMs). In the proposed scheme, first, SVM models are constructed using different kernel functions of linear, polynomial, radial basis, and sigmoid. The predictions of individual SVM models are then combined using majority voting. In this way, the combined decision space becomes more informative and discriminant. A comparative analysis of the proposed scheme is carried out with previous techniques. It is found that our scheme is more accurate for synthesized-blurred and real defocussed images.