The whole world is looking curious about future, anxious to understand what will happen in the next moment. Correspondingly, retailers are moreover curious about the destiny of their business, it's empowering and their future arrangements. Walmart is the world's greatest retailer and furthermore has a tremendous staple chain over the world. It was at first settled in America 1962. In 2019, it has in excess of 11,000 stores in 28 nations however the deals vary all around. Numerous business procedures, rebate rates will be presented for the improvement of deals. Retailers consistently attempt to pull in the everyday citizens to visit their store. They generally center on improving the future deals. Utilizing some Machine picking up determining models, we can evaluate the future deals dependent on the past information. Our point is to apply time arrangement gauging models to retail deals information, which contains week after week deals of 45 Walmart stores across United States from 2010 to 2012. There are different components which impacts the investigation of week after week deals - markdown, buyer per record, IsHoliday (Boolean worth returns whether it is occasion or not), size of the store, joblessness, store type, fuel cost and temperature. The anticipating models applied for the information are Auto Regressive Integrated Moving Average (ARIMA) model and Feed Forward Neural Networks (FFNN). The dataset will be separated into preparing and testing datasets. The anticipated qualities will be checked with the test information and precision will be determined. In view of the exactness we finish up which of the two models will better for the business expectation.