We have implemented the neural network model for the time series prediction of our sales.
Time series analysis serves as the basis for the demand forecast. In this type of model, tracking selected independent variables over a defined period enables forecasting predictable fluctuations in demand according to a prevailing trend, such as peak versus low periods.
We decided time series forecasting technique using neural network approach because of the following pros:
Easier to analyze segments
Most data is captured (historical data factors in the most important variables)
Reliance on own company history instead of multiple external data sources, which reduces the risk of breakdowns in data
Repeatability over days and weeks
How we used time series in our problem of sales forecast?
We framed time series as a supervised learning problem. Given a sequence of numbers for a time series dataset, we can restructure the data to look like a supervised learning problem. We can do this by using previous time steps as input variables and use the next time step as the output variable.
The use of prior time steps to predict the next time step is called the sliding window method.
For eg: we considered a window of size = 29 so that the sales data from the last 29 days and the present day are recorded as prior inputs and the sales of any day specified by the lag parameter can be calculated for e.g.:sales (t +90) i.e., sales forecast for a day 90 days ahead from present date.
——– by Sharadia Choudhury
Hadoop is an open-source framework of programs that is used to store and process big data. Hadoop uses multiple clusters of computers to analyze big data sets in parallel. The distributed processing of data sets can