Transactions on Machine Learning and Artificial Intelligence

Artificial God Optimization – A Creation

2020-01-08T17:24:34+00:00

Nature Inspired Optimization Algorithms have become popular for solving complex Optimization problems. Two most popular Global Optimization Algorithms are Genetic Algorithms (GA) and Particle Swarm Optimization (PSO). Of the two, PSO is very simple and many Research Scientists have used PSO to solve complex Optimization Problems. Hence PSO is chosen in this work. The primary focus of this paper is on imitating God who created the nature. Hence the term "Artificial God Optimization (AGO)" is coined in this paper. AGO is a new field which is invented in this work. A new Algorithm titled "God Particle Swarm Optimization (GoPSO)" is created and applied on various benchmark functions. The World's first Hybrid PSO Algorithm based on Artificial Gods is created in this work. GoPSO is a hybrid Algorithm which comes under AGO Field as well as PSO Field. Results obtained by PSO are compared with created GoPSO algorithm. A list of opportunities that are available in AGO field for Artificial Intelligence field experts are shown in this work.

Streambank Erosion Susceptibility Index and Flood-prone Area Mapping along the Karra River, Hetauda, Central Nepal Sub-Himalaya

2020-02-08T02:13:08+00:00

The Karra River, one of the major tributaries of the Rapati River, is the 5^th order stream that extends for about 21.91 km length in 92 sq. km. of watershed area. It is situated in the southern region of the Hetauda City, which is under the rapid development as a settlement and industrial area. The Karra River area is frequently impacted by streambank erosion and flooding during the heavy rainfall in monsoon due to loosely consolidated sedimentary terrain of the Upper Siwalik Subgroup and the unconsolidated Late Quaternary Deposits, which are vulnerable to erosion. Morpho-hydraulic parameters and stream cross-sectional characteristics and parameters of streambank erosion susceptibility index (SESI) rating system were assessed along the Karra River at 19 transects. The rating of the SESI are based on bank angle, bank height ratio, root depth ratio, root density, surface protection, bank materials and characteristics of stratification. The flood-prone area map was prepared based on the morpho-hydraulic parameters of the stream based on the maximum bankfull depth. The ER and W/D ratio were estimated to determine the affinity of flooding and lateral instability of the stream. Near Bank Stress Index and SESI dealt with streambank erosion potential were assessed to estimate the streambank erosion rate.

Optimization of Biogas Electrical Power Generation using Neuro-Fuzzy Controller

2020-01-08T17:24:32+00:00

Biogas electrical power generation is a renewable energy which originated from biological materials. The technology design and model power system that predict and control the generation of biogas Electrical production. This research paper develops a Neuro-fuzzy controller model for generation of Biogas power production. A Neuro-fuzzy controller is design to the Biogas power system in order to improve the power quality delivery to the load. The set of 27 rules are written for proper training of biogas electrical data in the neural network. The training is used to control signal of the Biogas Power output of the system. The output of Neural Network unit is given as input to the de-fuzzification unit and the linguistic variables are converted back into the crisp form. Therefore the algorithm was designed to decide power supply to the load as to improve the performance of the biogas system using MATLAB/SIMULINK and Neuro-fuzzy model was developed for easy input of the data. The result shows that biogas electrical power output increased by 4.39kw, which is 54.8% increase when Neuro-fuzzy controller is incorporated. The improvement in the system is due to the training of input parameters of the biogas generated. The result obtained shows that there is Real Power improvement in Biogas system when Neuro-fuzzy is incorporated in the system model

SOFT COMPUTING SYSTEM FOR THE DIAGNOSIS OF HORMONAL IMBALANCE

2020-01-08T17:24:30+00:00

Soft computing, as a science of modelling systems, applies techniques such as evolutionary computing, fuzzy logic, and their hybrids to solve real life problems. Soft computing techniques are quite tolerant to incomplete, imprecise, and uncertainty when dealing with complex situations. This study adopts a hybrid of genetic algorithm and fuzzy logic in diagnosing hormonal imbalance. Hormones are chemical messengers that are vital for growth, reproduction, and are essential for human existence. Hormones may sometimes not be balanced which is a medical condition that often go unnoticed and it’s quite difficult to be diagnosed by medical experts. Hormonal imbalance has several symptoms that could also be confused for other ailments. This proposed system serves as support for medical experts to improve the precision of diagnosis of hormonal imbalance. The study further demonstrates the effective hybridization of genetic algorithm and fuzzy logic in resolving human problems.