Modified K-nearest Neighbor Algorithm Research Articles

The Application of Modified K-Nearest Neighbor Algorithm for Classification of Groundwater Quality Based on Image Processing and pH, TDS, and Temperature Sensors

The limited availability of water in remote areas makes rural communities pay less attention to the water quality they use. Water quality analysis is needed to determine the level of groundwater quality used using the Modified K-Nearest Neighbor Algorithm to minimize exposure to a disease. The data used in this study was images combined with sensor data obtained from pH (Potential of Hydrogen), TDS (Total Dissolved Solids) sensors and Temperature Sensors. The test used the Weight voting value as the highest class majority determination and was evaluated using the K-Fold Cross Validation and Multi Class Confusion Matrix algorithms, obtaining the highest accuracy value of 78% at K-Fold = 2, K-Fold = 9, and K- Fold = 10. Meanwhile, the results of testing the effect of the K value obtained the highest accuracy value at K = 5 of 67.90% with a precision value of 0.32, 0.37 recall, and 0.33 F1-Score. From the results of the tests carried out, it can be concluded that most of the water conditions are suitable for use.

The Classification Status of River Water Quality in Riau Province Using Modified K-Nearest Neighbor Algorithm with STORET Modeling and Water Pollution Index

The Department of Environment and Forestry, Pollution and Environmental Damage Control Division, has an active role in monitoring water quality in Riau Province. The rivers that are still monitored and managed are Kampar River, Siak River and Indragiri River. Division of Environment Pollution calculates river quality status manually using Microsoft Excel, this is not maximally done since this important information should be processed quickly. Division of water pollution must determine the right calculation to get the results of the water quality status. Because of many calculation formulas set by the government, the commonly used method is the STORET method and the Pollution Index. So, in overcoming the problem of classification, the researcher proposes the use of learning methods that can predict or determine the status of water quality with classification techniques on data mining that is Modified K-Nearest Neighbor (MKNN) which is a modification of K-NN. The calculation of the MKNN algorithm produced the highest accuracy of 85.10% at K = 5 using STORET result data as training data. While, using the Pollution Index data results, the highest accuracy is 76.92% at K = 1. Based on the analysis with attribute analysis, the attributes that influence the determination of river water quality are BOD, COD, NH3, Fecal Coli and Total Coli. This result can be taken into consideration by the Division of Environmental Pollution in the process of overcoming and reducing pollutant overload that exceeds quality standards.

A risk-averse approach for the planning of a hybrid energy system with conventional hydropower

We present a risk-averse two-stage stochastic programming model for the planning of a hybrid energy system with conventional hydropower component. Using Conditional Value-at-Risk as our measure of risk-aversion, we take into consideration the dispersion of the random total cost arising due to uncertain streamflow amount. We propose an exact solution approach based on scenario decomposition to solve our large scale problem. We then present a case study for the Mediterranean Region in Turkey and generate scenarios using a modified k-nearest neighbor algorithm for bootstrapping the historical time series data of Manavgat River. The results of our computational study show how an optimal solution differs based on the degree of risk-aversion and demonstrate the computational power of our solution approach. Our algorithm is able to solve instances that cannot be solved by CPLEX, furthermore, CPLEX requires 5.84 times more computation time than our algorithm.

Modified K-nearest Neighbor Algorithm with Variant K Values

In Machine Learning K-nearest Neighbor is a renowned supervised learning method. The traditional KNN has the unlike requirement of specifying ‘K’ value in advance for all test samples. The earlier solutions of predicting ‘K’ values are mainly focused on finding optimal-k-values for all samples. The time complexity to obtain the optimal-k-values in the previous method is too high. In this paper, a Modified K-Nearest Neighbor algorithm with Variant K is proposed. The KNN algorithm is divided in the training and testing phase to find K value for every test sample. To get the optimal K value the data is trained for various K values with Min-Heap data structure of 2*K size. K values are decided based on the percentage of training data considered from every class. The Indian Classical Music is considered as a case study to classify it in different Ragas. The Pitch Class Distribution features are input to the proposed algorithm. It is observed that the use of Min-Heap has reduced the space complexity nonetheless Accuracy and F1-score for the proposed method are increased than traditional KNN algorithm as well as Support Vector Machine, Decision Tree Classifier for Self-Generated Dataset and Comp-Music Dataset.

Two-level hierarchical combination method for text classification

Text classification has been recognized as one of the key techniques in organizing digital data. The intuition that each algorithm has its bias data and build a high performance classifier via some combination of different algorithm is a long motivation. In this paper, we proposed a two-level hierarchical algorithm that systematically combines the strength of support vector machine (SVM) and k nearest neighbor (KNN) techniques based on variable precision rough sets (VPRS) to improve the precision of text classification. First, an extension of regular SVM named variable precision rough SVM (VPRSVM), which partitions the feature space into three kinds of approximation regions, is presented. Second, a modified KNN algorithm named restrictive k nearest neighbor (RKNN) is put forward to reclassify texts in boundary region effectively and efficiently. The proposed algorithm overcomes the drawbacks of sensitive to noises of SVM and low efficiency of KNN. Experimental results compared with traditional algorithms indicate that the proposed method can improve the overall performance significantly.