This project helps Walmart better predict sales of weather-sensitive products.
It seems obvious that the sales volume of products are influenced by many factors, such as holidays, prices, average per capita income, weathers and so on. Then, how can retailers matching their inventory levels with uncertain market demands to optimize their profits? Our group is interested in one of many factors that will affect the order quantity, weather.
The dataset that we are investigating comes from a Kaggle competition for Walmart weather-sensitive products’ sales prediction. The datasets was in csv formats for us to download. There are three datasets was given: a training dataset to train your algorithms, a testing dataset to test the accuracy of your prediction result, a key data set as a complementary information and a weather dataset as a record of the weather information during the period of time of our interest. In these datasets, the units sold of 110 items across 45 Walmart stores and covered by 20 weather station was given. The detailed weather measurements across the day in 2 years were recorded along with the sale of the item at that day.