RETAIL STORE DEMAND FORECASTING AND PREDICTION

Abstract

Forecasting is a crucial project in retailing. Especially, client-oriented markets which includes seasonality, trends, brief life cycles and a lack of ancient sales statistics which toughen the demanding situations of manufacturing correct forecasting. This overview paper offers trendy methods inside the income forecasting studies with a focal point on tendencies and new product forecasting. In this project we have explored different time series techniques on a relatively simple and clean dataset(collection of data). We had been given 5 years of retail store item sales data and asked to predict 3 months of sales for 5 different items at 10 different stores. We have conducted a top level view of ongoing advancement in the area of sales prediction & forecasting with the point of interest in developments & new item deals forecast. Traditional determining forecasting techniques face difficulties in generating correct income facts for brand spanking new merchandise and purchaser- orientated items. Specifically, uncertain call for, trends & seasonality, item variability just as a absence of historic can barely be dealt. In lately brought methods, ARIMA forecasting models perform more precisely. More specifically, we have a few years worth of daily sales data per product in each store, and goal is to forecast the future sales of each item in each store. Forecasting and prediction have been developed using dataset which had been used after training and testing of dataset. Also trend analysis and seasonal decomposition of a dataset(collection of data) for the purpose of checking if the dataset is authentic or not for retail's forecasting. At the end functions have been applied in the forecasting model(ARIMA and XGBoost) that’s how forecasting achieved. So far we’ve considered breaking down each product-store pair into a single time series, and doing a forecast for each time series, wise models for forecasting multiple time series collection in actual real-world systems. In other words, we used only the historical information of a particular store's sales of the product to forecast the future sales of that product in that store.