International Journal of Innovative Technology and Exploring Engineering (IJITEE) ISSN: 2278-3075, Volume-8 Issue-8 June, 2019 731 Published By: Blue Eyes Intelligence Engineering & Sciences Publication Retrieval Number H6749068819 /19©BEIESP Analysing various Regression Models for Data Processing Abstract: For modeling and analyzing several variables, many techniques are available among which in statistical modeling, regression analysis is one. Regression Analysis (RA) is utilized for prediction and determination, where its utilization has generous cover with the field of Artificial Intelligence. RA is a measurable procedure’s for assessing the relationship among variables (one dependent and one or more independent). Its helps us to predict and that is why it is also called as predictive analysis model. In this study, we had used vehicle data like velocity with which traffic move’s, gradient, actual velocity to predict the velocity profile of the vehicle. Also, we had analyzed various regression models like linear regression, multivariate linear regression and nonlinear regression. The outcome of this work is to write a function for every model that everyone can reuse that without using pre-defined functions in languages and plotting the given data to best fit for analyzing. Keywords; Regression, Predictor, Dependent variable, Machine learning, Vehicle and Velocity I. INTRODUCTION Regression Analysis (RA) is a group of factual devices that can help from numerous points of view to anticipate things of different segments. RA is utilized to construct numerical models to anticipate the estimation of one variable from learning of another. Figure 1 shows the eight different types of data mining techniques. Most regularly, RA evaluates the contingent desire for the needy variable given the free factors – that is, the normal estimation of the needy variable when the autonomous factors are fixed. Less usually, Revised Manuscript Received on June 05, 2019 Dr. K. K. Baseer, Associate Professor of IT & Member, Data Analytics Research Center, Sree Vidyanikethan Engg. College, Tirupati, India E- mail: drkkbaseer@gmail.com Vikram Neerugatti, Research Scholar, Department of CSE, SVUCE, Sri Venkateswara University, Tirupati, India. E-mail: vikramneerugatti@gmail.com Dr. Sandhya Tatekalva, Academic Consultant, Department of Computer Science, S.V. University, Tiruapti, India. E-mail: geetasandhya@gmail.com Dr. Akella Amarendra Babu, Department of Computer Science and Engineering, St. Martin’s Engineering College, Telangana, India E-mail: aababu.akella@gmail.com the attention is on a quantile, or other area parameter of the contingent circulation of the reliant variable given the autonomous factors. Figure 2 shows the machine learning types. In all cases, a regression function is used to assess the independent variables. In RA, it is additionally important to describe the variety of the needy variable around the expectation of the regression function using probability distribution. Figure 3 shows the three metrics used in regression. They are linear, logistic, exponential, nonlinear, polynomial, etc. commonly used regression is linear regression. In this we developed linear regression, multi variate linear regression, Gaussian regression using kernel, polynomial regression [1] [2] [3]. All these regression methods are done by using machine learning. We used Mat lab platform to solve regression analysis and developed various functions like gradient descent, cost compute, normalization. For experimental, vehicle data is used like velocity with which traffic move’s, gradient, actual velocity to predict the velocity profile of the vehicle. From these predicted range of a vehicle. By using different models of regression we come to conclusion which model predicts best and fits the data to the best. K. K. Baseer, Vikram Neerugatti, Sandhya Tatekalva, Akella Amarendra Babu S. No. Application Regression used to 1. Pharmaceutical company  Assess the stability of the active ingredient in a drug.  Predict its timeframe of realistic usability so as to meet FDA guidelines and  Identify a reasonable lapse date for the medication. 2. Credit Card company  Predict month to month gift voucher deals.  Improve yearly income projections . 3. Hotel Franchise  Identify a profile.  Predict potential customers. 4. Insurance company  Determine the probability of a genuine issue existing. Fig.1 Eight Data Mining Techniques (Courtesy: Google) TABLE 1: APPLICATIONS OF REGRESSION Fig. 2: Classification of Machine Learning (Courtesy: Google)