Artificial Intelligence Trends for Data Analytics Using Machine Learning and Deep Learning Approaches

K. Gayathri Devi, Mamata Rath, Nguyen Thi Dieu Linh, K. Gayathri Devi, Mamata Rath, Nguyen Thi Dieu Linh

• Editorial: CRC Press
• Año: 2020
• ISBN:
  9781000179538
• Edición: 1
• Categoría: Informatik
• Categoría: Computer Vision & Mustererkennung

Artificial Intelligence (AI), when incorporated with machine learning and deep learning algorithms, has a wide variety of applications today. This book focuses on the implementation of various elementary and advanced approaches in AI that can be used in various domains to solve real-time decision-making problems.

The book focuses on concepts and techniques used to run tasks in an automated manner. It discusses computational intelligence in the detection and diagnosis of clinical and biomedical images, covers the automation of a system through machine learning and deep learning approaches, presents data analytics and mining for decision-support applications, and includes case-based reasoning, natural language processing, computer vision, and AI approaches in real-time applications.

Academic scientists, researchers, and students in the various domains of computer science engineering, electronics and communication engineering, and information technology, as well as industrial engineers, biomedical engineers, and management, will find this book useful. By the end of this book, you will understand the fundamentals of AI. Various case studies will develop your adaptive thinking to solve real-time AI problems.

Features

• 
  
  

• Includes AI-based decision-making approaches

• 
  
  

• Discusses computational intelligence in the detection and diagnosis of clinical and biomedical images

• 
  
  

• Covers automation of systems through machine learning and deep learning approaches and its implications to the real world

• 
  
  

• Presents data analytics and mining for decision-support applications

• 
  
  

• Offers case-based reasoning

1

An Artificial Intelligence System Based Power Estimation Method for CMOS VLSI Circuits

Govindaraj Vellingiri

Sri Venkateswara College of Engineering and Technology

Ramesh Jayabalan

PSG College of Technology

Contents

• 1.1 Introduction
  • 1.1.1 Previous Work Using BPNN and ANFIS
• 1.2 Training and Testing Data
• 1.3 Power Estimation Using a Neural Network
  • 1.3.1 Construction of a Neural Network
  • 1.3.2 BPNN Training Phase
  • 1.3.3 BPNN Testing Phase
  • 1.3.4 Network Parameters
• 1.4 Proposed Power Estimation Using ANFIS Technique
  • 1.4.1 Overview of the Proposed Work
  • 1.4.2 Training and Checking Used in ANFIS
  • 1.4.3 Designing the ANFIS
• 1.5 Results and Discussions
  • 1.5.1 BPNN-Based Method
  • 1.5.2 Calculating Prediction Error
  • 1.5.3 ANFIS-Based Method
  • 1.5.4 Performance Evaluation
• 1.6 Conclusion
• References

1.1 Introduction

With the increased use of portable devices such as laptops, cellular phones, etc., power consumption has become a major issue that determines battery life span. Advances in very large scale integration (VLSI) technology have led to the fabrication of chips that contain millions of transistors. In nanometer or deep submicron technology, power consumption has become an essential concern due to factors such as the increase in number of transistors on a chip and speed due to scaling of transistor size. Hence there is a need to minimize power dissipation [1]. Under these constraints, it is necessary to estimate the average power consumption during the design phase. This chapter explains how to employ artificial intelligence systems such as back-propagation neural network (BPNN) and adaptive neuro-fuzzy inference system (ANFIS), which have the capability to accurately estimate power for the CMOS VLSI circuits, without knowledge of circuit structure and interconnections.

1.1.1 Previous Work Using BPNN and ANFIS

Application of ANFIS for computing resonant frequency in microstrip antennas was discussed in the literature [2]. Al-Shammari et al [3] designed an ANFIS system to estimate wind farm wake effect. Bhanja and Ranganathan [4] proposed a Bayesian network based method of switching activity measurement in VLSI circuits that employs logic-induced directed acyclic graphing within a realistic time and accuracy. Simulation and non-simulation-based approaches of average power were discussed in literatures [5] and [6]. Application of BPNN to estimate power for combinational and sequential circuits is discussed in references [1, 7–9]. Karimi et al [10] extracted small-signal equivalent circuit model (S-parameter data) of bipolar transistors using ANFIS. Guney et al [11] discussed resonant frequency calculation for microstrip antennas using ANFIS. Güler et al [12] presented a new approach using ANFIS for coplanar waveguides gap discontinuities identification. Fault classification for PLL using BPNN was presented by Ramesh et al [13]. Application of ANFIS to predict air temperature to provide knowledge about climate and drought detection was discussed by Karthika et al [14], in which the authors compared and showed that the Gaussian membership function performs better than the Gbell membership in ANFIS. Estimation of power using BPNN that used the Levenberg-Marquardt function was proposed by Hou et al [15]. Mohammadi et al [16] estimated wind-power density by using a model based upon extreme learning machine (ELM), showing that BPNN can also be used for automating power estimation. Muragavel et al [17] discussed estimating average power through exhaustive simulation for larger input circuits. Nikolić et al [18] demonstrated application of ELM for sensor-less wind-speed predictions. Nikolić et al [19] explained wake power and wind speed deficit prediction using soft computing techniques. Nikolić et al [20] proposed an application of ANFIS to estimate wind turbine noise levels. Nikolić et al [21] discussed statistical analysis such as root mean square error (RMSE) and coefficient of determination (R) of wind speed using ANFIS. Applying ANFIS to wind-power modeling and wind turbines were discussed in reference [22]. ANFIS-based prediction of the modulation transfer function of an ...

Índice

• Cover
• Half Title
• Series Page
• Title Page
• Copyright Page
• Contents
• Preface
• Acknowledgment
• Editors
• Contributors
• Chapter 1 An Artificial Intelligence System Based Power Estimation Method for CMOS VLSI Circuits
• Chapter 2 Awareness Alert and Information Analysis in Social Media Networking Using Usage Analysis and Negotiable Approach
• Chapter 3 Object Detection and Tracking in Video Using Deep Learning Techniques: A Review
• Chapter 4 Fuzzy MCDM: Application in Disease Risk and Prediction
• Chapter 5 Deep Learning Approach to Predict and Grade Glaucoma from Fundus Images through Constitutional Neural Networks
• Chapter 6 A Novel Method for Securing Cognitive Radio Communication Network Using the Machine Learning Schemes and a Rule Based Approaches
• Chapter 7 Detection of Retinopathy of Prematurity Using Convolution Neural Network
• Chapter 8 Impact of Technology on Human Resource Information System and Achieving Business Intelligence in Organizations
• Chapter 9 Proficient Prediction of Acute Lymphoblastic Leukemia Using Machine Learning Algorithm
• Chapter 10 Role of Machine Learning in Social Area Networks
• Chapter 11 Breast Cancer and Machine Learning: Interactive Breast Cancer Prediction Using Naive Bayes Algorithm
• Chapter 12 Deep Networks and Deep Learning Algorithms
• Chapter 13 Machine Learning for Big Data Analytics, Interactive and Reinforcement
• Chapter 14 Fish Farm Monitoring System Using IoT and Machine Learning
• Index