PREDICTIVE ANALYSIS FOR BIGMART SALES USING MACHINE LEARNING

This the streamlit page which appears after executing the program
This the predicted output which displays after clicking on the button predict

Inspiration

A sales forecast combines both art and science for development. Accurate forecasts of revenue keep administration content, the business expanding and the marketing staff motivated. Making informed, deliberate sales decisions will be facilitated by having a general idea of what to anticipate because sales are the backbone of any company structure.

What it does

Big Mart sales predictions can be made with accuracy by using machine learning algorithms. To find patterns and trends, these algorithms can evaluate historical sales data, demographic data, product features, and other pertinent aspects. Companies can obtain insights into client preferences and behavior by utilizing machine learning algorithms, which enables them to decide wisely about inventory management, pricing schemes, and marketing efforts. Additionally, machine learning can assist in e-commerce sectors in identifying anomalies and fraudulent activity, ensuring that clients have a safe and reliable buying experience.

How we built it

To forecast sales, we used the Random Forest Regressor and the Linear Regression. The train_test_split() function was used to construct a validation set. Test size should be set to 0.25 so that 75% of the data points are in the train set and 25% are in the validation set.

Challenges we ran into

We have faced challenges when we were pre processing the data as we had anomalies in the data

Accomplishments that we're proud of

We have stepped into the data science field and started this project with great enthusiasm. We are proud that we have completed our project with less RMSE value.

What we learned

Hypothesis Generation – understanding the problem better by brainstorming possible factors that can impact the outcome. Data Exploration – looking at categorical and continuous feature summaries and making inferences about the data. Data Cleaning – imputing missing values in the data and checking for outliers. Feature Engineering – modifying existing variables and creating new ones for analysis. Model Building – making predictive models on the data.

What's next for PREDICTIVE ANALYSIS FOR BIGMART SALES USING MACHINE LEARNING

We can use the various machine- learning algorithms, such as linear regression and decision tree algorithms, and an XGBoost regressor, which offers an efficient prevision of Big Mart sales based on gradient.