255-FinalProject

Team 15 Walmart Sales Prediction [ Algorithm Track ]

• Vamshidhar Reddy Parupally - 016001427
• Tirupati Venkata Sri Sai Rama Raju Penmatsa - 016037047

Details

• Dataset : https://www.kaggle.com/competitions/walmart-recruiting-store-sales-forecasting

• Colab notebook for Vamshidhar Reddy's Contribution [also included file in github]: https://colab.research.google.com/drive/1PZpK60do4ptjMqOB4-XqOD75bDw2ygNl?usp=sharing

• Colab notebook for Tirupati Venkata Sri Sai Rama Raju's Contribution : https://colab.research.google.com/drive/1fGAk9V3JTfaEljtP4_vEmfswIwGbq0gA?usp=sharing

• Video Demo : https://drive.google.com/drive/folders/1Y_5N2be5BYA1S2c6qlq-oiWD9VdRFNs7?usp=sharing

Contributors

Vamshidhar Reddy Parupally - Contribution

---

Tirupati Venkata Sri Sai Rama Raju Penmatsa - Contribution

---

About Dataset

• This project comprises a thorough data analysis, as well as time series analysis and sales forecasting using multiple models.�
• The data was collected between 2010 and 2012, and 45 Walmart locations across the country were examined.

Dataset Description

1. Stores:

• Store: store numbers rangingfrom 1–45.
• Type: store type ‘A’, ‘B’ or ‘C’.
• Size: no. of products available in the particular store ranging from 34,000 to 210,000

2. Sales:

• Date: The date of the week.
• WeeklySales: sales during that Week.
• Store: The store number 1–45.
• Dept: One of 1–99.
• IsHoliday: Boolean value representing a holiday week or not.

3. Features:

• Temperature: Temperature of the region during that week.
• FuelPrice: Fuel Price in that region during that week.
• MarkDown1:5 : Represents the Type of markdown and what quantity was available during that week.
• CPI: Consumer Price Index during that week.
• Unemployment: The unemployment rate during that week in the region of the store

Approach 1 and Approach 2 Pre Processing:

• Merged the 3 different csv files to form the actual dataset to make the training easy.

• Correlation analysis,

• 

• Null value analysis,

• 

• Outlier analysis,

• 

• Average weekly sales analysis,

• Label encoding

• Marking the NA values as 0 for “MarkDown” ‘s and dropping “CPI” and “Unemployment”
• Removing Anomalies

Algorithms used:

LSTM (state of the art)

Light GBM (Boosting)

XG Boost (Boosting)

ANN [ 5 Different Configurations ]

Random Forest Regressor

Decision Tree

ExtraTreeRegressor

Application of PCA on the data

Comparison of the models

Approach 1

Approach 2

References

• [1] https://www.kaggle.com/code/bhatnagardaksh/walmart-sales-predictionRandom-Forest
• [2] https://medium.datadriveninvestor.com/walmart-sales-data-analysis-sales-prediction-using-multiple-linear-regression-in-r-programming-adb14afd56fb
• [3] https://www.kaggle.com/competitions/walmart-recruiting-store-sales-forecasting/code
• [4] https://www.kaggle.com/competitions/walmart-recruiting-store-sales-forecasting
• [5] https://www.kaggle.com/code/datamany/random-forest-rnn-walmart-sales-forecastMachine-Learning
• [6] https://machinelearningmastery.com/xgboost-for-regression/
• [7] https://www.analyticsvidhya.com/blog/2021/03/introduction-to-long-short-term-memory-lstm/
• [8] https://www.superdatascience.com/blogs/recurrent-neural-networks-rnn-the-vanishing-gradient-problem
• [9] https://www.analyticsvidhya.com/blog/2017/06/which-algorithm-takes-the-crown-light-gbm-vs-xgboost/
• [10] https://www.geeksforgeeks.org/ml-label-encoding-of-datasets-in-python/
• [11] https://towardsdatascience.com/nlp-101-word2vec-skip-gram-and-cbow-93512ee24314
• [12] https://towardsdatascience.com/time-series-data-analysis-resample-1ff2224edec9
• [13] https://towardsdatascience.com/the-complete-guide-to-time-series-analysis-and-forecasting-70d476bfe775/
• [14] https://towardsdatascience.com/efficient-time-series-using-pythons-pmdarima-library-f6825407b7f0/
• [15] https://www.analyticsvidhya.com/blog/2018/08/auto-arima-time-series-modeling-python-r/