STORE SALES FORECASTING

TEAM 20
Abhijeet Jawlekar
Akash Meena
Lubhawan Parihar
Keshav Basandrai
Vivek Jain

INTRODUCTION
Forecasts aren’t just for meteorologists. Governments forecast economic growth. Scientists
attempt to predict the future population. And businesses forecast product demand—a common
task of professional data scientists. Forecasts are especially relevant to brick-and-mortar grocery
stores, which must dance delicately with how much inventory to buy. Predict a little over, and
grocers are stuck with overstocked, perishable goods. Guess a little under, and popular items
quickly sell out, leading to lost revenue and upset customers. More accurate forecasting, thanks
to machine learning, could help ensure retailers please customers by having just enough of the
right products at the right time.
Current subjective forecasting methods for retail have little data to back them up and are unlikely
to be automated. The problem becomes even more complex as retailers add new locations with
unique needs, new products, ever-transitioning seasonal tastes, and unpredictable product
marketing.

METHODS
1) Data
i) Train data

(1) The training data, comprising time series of features store_nbr, family,

and onpromotion as well as the target sales.
(2) store_nbr identifies the store at which the products are sold.
(3) family identifies the type of product sold.
(4) sales gives the total sales for a product family at a particular store at a given date.
Fractional values are possible since products can be sold in fractional units (1.5 kg of
cheese, for instance, as opposed to 1 bag of chips).
(5) onpromotion gives the total number of items in a product family that were being
promoted at a store at a given date.

ii) Stores data

(1) Store metadata, including city, state, type, and cluster.
(2) cluster is a grouping of similar stores.

iii) Oil data

 (1) Daily oil price. Includes values during both the train and test data timeframes.
(Ecuador is an oil-dependent country and it's economical health is highly vulnerable
to shocks in oil prices.)
iv) Holidays data
(1) Holidays and Events, with metadata
(2) Additional holidays are days added a regular calendar holiday, for example, as
typically happens around Christmas (making Christmas Eve a holiday).

2) EDA (Exploratory data analysis)

a) Analyzed sales data weekly, monthly and yearly using seaborn library.
b) Using Pearson correlation coefficient analyzed correlation between total no of on
promotion items with Total sales.
c) Computed avg sales by family product, to find highest no of product being sold.
d) Analyzed sales per store averaged over total time period.
e) Determined seasonality in the data using seaborn and periodogram.
f) Plotted lag plots and partial auto correlation function to find out importance of different
lags for current time step.

3) Models Used:
a) SARIMA

b) LSTM

c) BiLSTM

d) TBATS
 PROJECT TITLE
TEAM NAME: Give your team a fancy name
TEAM NUMBER: Which team number?
TEAM MEMBERS: All members of your team (should match the spreadsheet)

Introduction (Minimum 100 words)

Provide a brief introduction to the problem, in your own words.

Methods (Minimum 300 words)

Provide a brief description of your approach. Draw pictures if needed.

Results (Minimum 400 words)

Describe the salient results. Provide enough details so that the conclusions can describe insights,
novelty, etc. Draw pictures if needed.

Conclusion (Minimum 200 words)

What are your main conclusions?

Evaluation criteria (max 30)

You have 12 minutes for your presentation.

Criterion Marks
Definition of scope, clarity of the presentation of scope 5
(Did you describe the problem clearly? Did you define the scope and
goal?)
Solution methodology 5
(Did you describe your proposed approach clearly?)
Description of results 5
(Did you describe your results crisply, clearly, and completely? Did you
meet the goals you set for your team?)
Innovation, new insights, discussion 5
(Did you do something new? Did you create an ‘Aha’ moment for your
audience?)
Response to questions 5
(How good were your responses to the questions)
Effort, substantiality, and technical depth 5
(How much did you achieve? Was it a hotch-potch, previous night-out
effort, or did you plan and systematically make progress in your project
(as evidenced by your presentation)?)