Random forest (RF) selects feature subsets randomly. Useless and redundant features will lower the quality of selected subsequently affect overall classification accuracy RF. This study proposes an improved RF algorithm based on hierarchical clustering (HCRF). The uses algorithms to optimize selection process, by establishing similar groups GINI index, then selecting from each group proportionally construct subset. subset is used a single classifier. process increases filtering subsets,...

Intelligent Tutoring System (ITS) is one of the solutions due to need for on-demand tutoring among students nowadays. ITS may provide unlimited access effective and affordable personal anytime, anywhere. It developed specially give a one-to-one while simulating student-teacher learning environment. The process includes giving notes, examples, exercises, hints corrections, similar in-class process. main objective this study systematic view implementing two different artificial intelligence...

The distribution of data plays an important role in determining the successfulness learning process machine learning. Data sets with imbalanced may lead to biased results, especially clustering. If is insufficient, clustering will not be able cluster and this add randomness grouping. Therefore, KSOM algorithm modified improve process. This modification done based on exploration exploitation procedures Ant Clustering Algorithm (ACA). To investigate effectiveness algorithm, three are chosen;...

The Kohonen Self-Organizing Map (KSOM) is one of the Neural Network unsupervised learning algorithms. This algorithm used in solving problems various areas, especially clustering complex data sets. Despite its advantages, KSOM has a few drawbacks; such as overlapped cluster and non-linear separable problems. Therefore, this paper proposes modified that inspired from pheromone approach Ant Colony Optimization. modification focusing on distance calculation amongst objects. proposed been tested...

It is crucial to differentiate between venomous and non-venomous snake whereby an immediate effective medical care can be instituted the victims. However, early identification not easy. We developed a bite system (N'viteR) using Neuro-GA technique. Based on 200 cases, this work had revealed that has yield high accuracy in identifying snake. A number of experiments have been done which based epoch, momentum, learning rate, generation, population chromosome. This hybrid technique achieved...

Tropical wood recognition is a very challenging task due to the lack of discriminative features among some species wood, and also inter class species. Moreover, noises illuminations, or uncontrolled environment as well such size pores, density etc., which depend much on age, weather other factors, contributing irregularities features. In this paper, we explore use feature extraction techniques, classification techniques for better accuracy system. particular, one deep learning method...

The Ant Clustering Algorithm (ACA) is a biological inspired data clustering technique, which aimed to cluster and classify the patterns into different groups. This paper shows how implemented in classifying tropical wood species. As for feature extraction this research, two extractors are selected extract features from images, Basic Grey Level Aura Matrices (BGLAM) Statistical Properties of Pores Distribution (SPPD). ACA algorithm then been applied training testing, as result, it proven that...

Identifying types of snakes is crucial for appropriate anti-venom administration. We developed a Snake Bite Diagnosing System based on standard Back Propagation and Resilient Neural Networks. These systems were capable in differentiating between venomous non-venomous snakes. The accuracy both analyzed compared. post development comparative studies revealed that the technique yielded 83.33%, 90.00% respectively mean squared error, number epoch parameter setting. Whereas Standard produced...

Clustering is proposed to cluster the high dimensional data, into clusters of data that exhibit some similarities. Due this ability, it has been chosen solve many problems in various areas, including tropical wood species classification. It ease recognition process which done manually before. Pheromone-based Kohonen Self-Organizing Map (PKSOM) a clustering tool datasets; filtered and raw datasets. This paper discusses performance scalability modified (KSOM), named KSOM (PKSOM). The PKSOM...

This research is addresses to determine the dominant species that located in overlapped clusters produced by Kohonen Self-Organizing Map (KSOM). Before, KSOM algorithm able cluster tropical wood data set effectively and accurately according features, which pores sizes. Unfortunately, there are seven clustering result this due similarity features among species. problem has caused difficulty determining separation boundary amongst clusters, where most for every difficult identify. As a...

The assessment of water microbial quality is normally performed by verification ofEscherichia coli where the growth in nonlinearity. NARX computational tools that haveextensive utilization solving nonlinear time series problems. It well known as one thetechnique has ability to predict with efficient and good performance. Using NARX, ahighly accurate model was developed Escherichia (E. coli) basedon pH parameter. multiparameter portable sensor spectrophotometer data wereused build train...

Travel agencies set new prices on travel packages based their experiences by analyzing the trend holiday and festive season. However, they find it hard to predict exact with minimum be offered for upcoming years. Prices keep changing due other reasons rather than This research paper applied data analytics which is divided into two parts, 1) descriptive facilitate have better insights of 2) predictive price forecasting. Visualization a part where dispersion correlation are produced gain...

A key to wood identification is the distinguishable features found on cross-sectional surface of each tree species. The pattern cross-section may look very similar non-experts. However, trained experts identify species based distinct and discriminant pattern. An automatic recognition system machine vision emulate experts, KenalKayu has been developed with high classification accuracy. Unfortunately, when more were added into system's database, accuracy reduced. It important for have a...