International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 08 Issue: 05 | May 2021 www.irjet.net p-ISSN: 2395-0072 © 2021, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1100 Telecom Churn Prediction Model using XgBoost Classifier and Logistic Regression Algorithm Miss.Priyanka Parmar 1 P.G. Scholar, MASTRE OF CE KITRC, Ahmedabad, Gujarat, India Mrs. Shilpa Serasiya 2 H.O.D, CE Department. KITRC, Ahmedabad, Gujarat, India --------------------------------------------------------------------***----------------------------------------------------------------------- Abstract:- We will increase gadget getting to know version which clients who've now no longer completed any usage, both incoming or outgoing - in terms of calls, net etc. over a length of time. A ability purchaser has stopped the use of the offerings for a while; it is able to be too overdue to take any corrective moves to keep them. For e.g., in case you outline churn primarily based totally on a ‘two-months 0 usage’ length, predicting churn may be vain when you consider that via way of means of that point the purchaser could have already switched to every other operator. For that set of rules we will use dataset carries purchaser-degree records for a span of four consecutive months - June, July, August and September. The months are encoded as 6, 7, eight and 9, respectively. We can be going to use logistic set of rules to discover climate purchaser is churn or now no longer. To discover a purpose why it's miles churn or now no longer we are able to follow XGBOOST Classifier set of rules to discover which function has affected to grow to be a churn or now no longer. We will first follow facts pre-processing and cleaning the facts the use of pandas and numpy library and carry out EDA (Exploratory facts analysis) Task. After, efficaciously completed EDA Analysis we are able to put into effect Logistic set of rules for prediction version and get higher overall performance and function significance we are able to put into effect XGBOOST Classifier which offer us capabilities which affected to purchaser grow to be a churn or now no longer. For in addition enhancement we are able to do Customer Behavioral Analysis the use of supervised strategies and acquire green facts on big scale. Keywords: Telecom Churn, EDA (Exploratory Data Analysis Xgboost (Extreme Gradient Boosting) Classification Algorithms. 1. INTRODUCTION Simple terms, customer churn occurs when the consumer wants to completely stop your services and switch to another provider. Customer churn has a major impact on businesses that rely primarily on subscription revenue. The amount of persons who leave a group over time is referred to it as the churn rate. Today's telecommunications firms are up against a lot of stress. competition today, as every corporation comes with a brand new scheme each month to draw clients. Losing clients in this type of enterprise may be very costly, as it's miles extraordinarily tough to draw NEW clients. Bad carrier opinions on the internet, phrase of mouth opinions are elements that discourage ability clients from becoming a member of a brand new network. It turns into extraordinarily crucial for such corporations, to hold its cutting-edge clientele. Finding out what may be viable elements that reason a purchaser to churn could assist corporations provide higher services custom designed to a specific consumer’s needs. At instances it is able to manifest that a precise consumer has a better want for records and the cutting-edge issuer turns out to be too steeply-priced and the consumer switches to any other corporation. Predicting which customers should viable transfer can assist telecom corporations devise a brand new plan for the ones unique clients so one can save you them from leaving the network. In reality this isn't always new, the instant you install a request to stop services, and also you get a name from them with a few new plan to trap you. However, from a non-public experience, as soon as a purchaser is bored stiff together along with your services, and has already made the selection to transfer, it's miles extraordinarily tough to persuade them to stay. It is higher to provide such plans BEFORE they make the selection. Supervised Machine Learning is not anything however gaining knowledge of a characteristic that maps an enter to an output primarily based totally on instance enter-output pairs. A supervised gadget gaining knowledge of set of rules analyzes the schooling statistics and produces an inferred characteristic, which may be used for mapping new examples. Given that we've got statistics on contemporary and previous purchaser transactions within side the telecom dataset, that is a standardized supervised category hassle that attempts to are expecting a binary outcome (Y/N).