ICT for Competitive Strategies

Proceedings of 4th International Conference on Information and Communication Technology for Competitive Strategies (ICTCS 2019), December 13th-14th, 2019, Udaipur, India

Durgesh Kumar Mishra, Nilanjan Dey, Bharat Singh Deora, Amit Joshi, Durgesh Kumar Mishra, Nilanjan Dey, Bharat Singh Deora, Amit Joshi

• Publisher: CRC Press
• Year: 2020
• ISBN:
  9781000191332
• Edition: 1
• Topic: Ciencia de la computación
• Subtopic: Inteligencia artificial (IA) y semántica

Fourth International Conference on Information and Communication Technology for Competitive Strategies targets state-of-the-art as well as emerging topics pertaining to information and communication technologies (ICTs) and effective strategies for its implementation for engineering and intelligent applications.

EXPLORATION OF SUPERVISED MACHINE LEARNING ALGORITHMS ON BINARY CLASSIFICATION

Dr. Anjali Jivani, Dr. Hetal Bhavsar, Sneh Shah and Riya Shah

Department of Computer Science and Engineering, The Maharaja Sayajirao University of Baroda, Vadodara, India

[email protected], [email protected], [email protected], [email protected]

ABSTRACT: Supervised learning, in the milieu of Machine learning, is a type of arrangement in which we have known data with their predefined classes which help in designing classifiers. These classifiers are the base of learning for processing the data in future. This paper presents the empirical survey conducted on different Supervised Learning algorithms on varied datasets. The detailed analysis on which supervised method proves to be more efficient on what kind of data is carried out and discussed. The implications on the number and type of features are also discussed. The performance metrics used are the Accuracy, Precision and F1-score.

Keywords: Machine Learning, Supervised Learning, Data Mining Techniques, Data Analysis, Classification

I. INTRODUCTION

Machine learning (ML) is alucrative branch or implementation of artificial intelligence (AI). It gives systems the ability to learn involuntarily. It improves with practice as it not programmed explicitly. The basic hyposthesis of machine learning is to construct algorithms that receives input in the form of datasets and uses statistical analysis to predict an output. As the new data is available, it keeps on updating the output from datasets. It comes under AI as the main aim is to allow computers to keep on learning and improving based on the new and varied inputs. It is broadly categorized in two parts – supervised and unsupervised learning.

Supervised machine learning algorithms as the name suggests, uses data with predefined labels to predict the labels of the new data. Using the learning algorithm to produce a function to make predictions about output values, the first step begins with the analysis of a known training dataset. This is done by training the classifier first and then actually implementing it on new unseen data. To modify the model according to the requirements, the algorithm compares its output with correct output and rectifies errors. Supervised learning is popularly known as Classification or Categorization techniques.

In contrast, unsupervised machine learning algorithms are implemented on totally unseenand previously unknown data. The unsupervised learning techniques group or cluster these data in such a way that within the clusters they are very similar in their feature values whereas across clusters they differ. Unsupervised learning is prevalently known as Clustering.

The work conducted here is focused on supervised machine learning algorithms considering only binary classification i.e. two classes only. The datasets used for this work has been taken from the UCI Machine Learning Repository [18]. The methods considered in this study are some of the popular machine learning algorithms like KNN, DECISION TREE, SVM, RANDOM FOREST, NEURAL NETWORK, GAUSSIAN NAÏVE BAYES, GAUSSIAN PROCESS CLASSIFIER and ADABOOST. To analyze the working of these classification algorithms in terms of efficiency and accuracy, the metrics used are accuracy, precision, recall and F1-score. The method to apply supervised machine learning techniques to a real world problem is delineated in figure below:

Figure 63.1 Predictive stages of sup...

Table of contents

• Cover
• Title Page
• Copyright Page
• Table of Contents
• Lists of Figures
• List of Tables
• Preface
• ICTCS 2019 Committee Members
• Technical Program Committee
• Internet of Things (IoT) – Enabler for Connecting World
• Hardware Based Data Compression Using Lempel-ZIV-WELCH Algorithm
• Detection of Eye Diseases using Image Processing and Artificial Neural Networks
• A Hybrid Image Watermarking Methodology Based on BP-RBF Neural Network Model and Sift Transform Function
• Applying Data Science Solutions in Healthcare Industry
• Prediction of Weblogs for Improving the Performance in Web Usage Mining using Psosvm
• Comparative Study of Machine Learning Algorithms on Sentiment Analysis of Product Reviews
• Convolutional Neural Network-Based Pattern Recognition Technique for Diabetic Retinopathy Detection
• Gate Engineered Vertical Mosfet with Reduced Specific Resistance and Switching Delay
• Smart Vacuum Robot
• A Hybrid Approach for Detection of Forgery in Video
• Design and Development of an Empirical Model to Predict Cognitive Learning Level Using Bloom Taxonomy for Undergraduate Engineering Students
• Semantic Framework and Methodology for Cultural Heritage Data Integration Fore-Walkthrough
• Design and Analysis of Folded Printed Quadrifilar Helical Antenna for Gps Application
• Application of Deep Learning of Rain Water Harvesting and Recycling in That Pure Water To Live With Photovoltaic System
• Improvised Curvelet Transform Based Diffusion Filtering for Speckle Noise Removal In Real-Time Vision-Based Database
• India Rankings: Impact on Research Publications (A Case Study on Top 20 Engineering Institutions)
• Optimized Singular Value Decomposition(SVD) Based Watermarking of Iris Biometric Data to Mitigate Replay Attack
• BI-Iterative Optimized K-Nearest Neighbours Algorithm
• A Review: Control Strategies for Power Quality Improvement of Hybrid Stand-Alone Solar Pv And Wind Energy System
• Multi-Controllers Based Subnet in SDN With BRS
• Categorization of Virtual Machine in Cloud SDN Environment using ELM – A Discriminative Classifier
• Definet: Portable CNN Network for Facial Expression Recognition
• Pre-Processing of Dogri Text Corpus
• A Novel DASR Approach For Reconstruction of Hyper-Spectral Images
• Identification of Genuine Images Out of Near Original and Replicas to Enhance The Machine Learning By Convolutional Neural Network
• HLReCapNet: Convnet to Detect High and Low-Resolution Screen Captured Images
• Internet of Things Based Embedded System for Smart Irrigation
• Noise Removal as Pre Processing Task and Its Implementation for Gujarati Named Entity Recognition
• LIDAR & Camera Based Assistance System for the Visually Challenged
• Futuristic Robotic Artificial Intelligence Systems for Special Needs Children
• Outlier Detection and Removal: An Efficient and Effective Concept in Healthcare Sector
• Quantification of Tool Wear Condition From the Sound Recorded During the Process Using Neural Network
• Provisioning of Broadband Communication for Passengers in Hyperloop using 5G Networks
• Performance Analysis of Polar Codes
• Identification of Leaf Disease using Improved Region Growing Method
• A Survey on Data Level Techniques – A Customer Churn Prediction Case Study
• A Study of Crime in India Through Statistical Analysis
• User Centred System Design for Indian Railway Parcel Service
• A Look at Top 35 Problems in the Computer Science Field for the Next Decade
• A Case Study of Enterprise Machine Learning Framework for Investment Platforms
• An Enhanced and Dynamic Crowd Estimation System
• Fog Computing for Internet of Medical Things Based Healthcare Systems
• Improving Performance of Collaborative Filtering
• Animate Object Detection and Q Ground Control
• Future Computing Technology: Quantum Computing and Its Growth
• Securing Medical Data Using Fully Homomorphic Encryption
• Cammod: Convolutional Neural Network Based Camera Model Identification
• A Hybrid Metaheuristic Approach for Model Based Testing of Object Oriented Programs
• Securing Web Applications using Security Patterns
• Design and Development of Low-Cost Portable System for Detection of Eye Diseases
• Learning Objects of Computer Science Discipline: An Analysis of E-PG Pathshala With Ugc-Net A Study for Future Sustainability with Emerging Topics
• Automated Person Authentication using Face, Iris and Ear Multimodal Biometric Fusion
• Thought to Text Conversion using Deep Learning
• Comparison of Regression Techniques for Predicting the Academic Performance of Students in Educational Data Mining
• Feature Evaluation for Learning Underlying Data-Processing to Enhance Cloud Trust Through Rich Models
• Fractional Delay Finite Impulse Response Fifilter Synthesis using Dierential Search Algorithm
• Analysis of PTS-MIMO-OFDM Signal for Goppa Coded Data
• Intelligent Cryptography Approach on Identity Based Encryption (IBE) for Secured Distributed Ehr Data Storage in Cloud Computing
• Encryption and Decryption Analysis of Non - Stationary Image using Canonical Transforms with Scrambling Technique
• Energy Efficient Strategies in Internet of Things: An Overview
• Pricing Based Caching Strategy in Named Data Networking
• Exploration of Supervised Machine Learning Algorithms on Binary Classification
• Survey Paper on Efficient Security Mechanism in Iot using Blockchain
• Comparative Analysis and Machine Learning Based Decision Making Financial Exchange Value Prediction Methodology in National Stock Exchange
• An Analysis on the Effectiveness of Utilizing Facebook Graph Structure for Trust Management In A Social Iot Network
• Net Neutrality for Digital Dependent Indian Business Scenario
• Intrusion Detection in Internet of Things Based E-Healthcare System - A Systematic Review
• Intelligent Wireless Multidirectional Fire Extinguishing Robot
• Contribution of Various Components in Cloud Trust: A Cloud Consumer’s Perspective
• Comparative Assessment of Machine-Learning Based Methodologies and Algorithms with Accuracy, Sensitivity and Specificity for Prediction of Heart Disease
• Best Practices in Higher Educational Institutions: Subjective Evaluation of PoS and Blooms Taxonomy
• Heart Disease Prediction Using Deep Belief Network Optimized by Particle Swarm Optimization
• One Step Technological Solution for Agriculture
• Unsupervised Learning Methods as Tools for Discovering Relationships within Data
• Cluster Based Load Balancing in Cloud Environment
• Blockchain in Agile Software Development
• Survey of Osteoporotic Bone Detection using Texture Analysis
• The Cyber Security Challenges: A Survey of Chief Information Security Officer in Indian Context
• Analysis of Convolutional Neural Network using Pre-Trained Squeezenet Model for Classification of Thermal Fruit Images
• Empty Region Detection in an Image using Faster R-CNN Architecture
• Dual Field ECC Processor for IOT Applications
• Comparative Analysis of Cloud Security Complexities and Past Proposed Non-Homomorphic and Homomorphic Encryption Methodologies with Limitations
• Multipoint Deflection Analysis using Non-Contact Image Based System
• Using Microscopic Images to Predict Plant Diseases in a Deep Learning Environment
• An Advanced Ensemble Approach to Household Poverty Level Prediction
• Modern Error Controlling Code for 5G: LDPC Codes
• Server-Less Intensive Vital Monitoring System (I-VMS)
• Predicting the International Traveler’s Visit to India and the Impact of Social Media on Their Visit using Logistic Regression
• Secured Decentralized Archiving Healthcare Data using Blockchain with IOT
• Analysis of Channel Characteristics in Wireless Underground Sensor Network
• Comparative Analysis of Switched Reluctance Motor and Bldc Motor for Electric Vehicles Using Ansys Maxwell
• Review of Clustering Solutions for ECG Heartbeat Arrhythmia Detection
• Multi Data Users Using Novel Privacy Preserving Protocol in Cloud Networks
• Assessment of Natural Vegetation Through NDVI Approach of Nanded City, Maharashtra, India
• Subject Index