Team 118

Inspiration

Our team was highly interested in the challenges faced by businesses and accurately predicting future sales with our data science knowledge. Unpredictable market dynamics, changing consumer behaviour, and external factors make it crucial for companies to have a robust sales forecasting model.

What it does

The Sales Forecasting Excellence model utilises machine learning algorithm Gradient Boosting Regressor to analyse historical domestic sales data and identify trends and patterns, such as correlations between different variables. It provides companies with accurate sales forecasts, enabling them to make informed strategic decisions, optimise inventory, and allocate resources effectively.

How we built it

We used Python library numPy for data preprocessing and analysis, scikit-learn for machine learning, and matplotlib/seaborn for data visualisation. The model went through rigorous training and validation using cross-validation techniques.

Challenges we ran into

One of the main challenges was dealing with extremely large inputs in the sales data, which required careful preprocessing. Meanwhile, selecting variables with relatively high correlations with the sales figures requires careful consideration and calculation. Additionally, selecting the most suitable machine learning algorithm posed a challenge, and we experimented with different models before finding the optimal solution.

Accomplishments that we're proud of

Successfully developing a model that accurately predicts sales figures and overcoming challenges related to data preprocessing and model selection.

What we learned

Our team gained valuable insights into the complexities of sales forecasting and the importance of robust data preprocessing. We enhanced our skills in machine learning and improved our ability to troubleshoot and optimize models.

What's next for Team 118

In the future, we plan to refine our model further by incorporating more advanced features and exploring the integration of external data sources. We also aim to develop a user-friendly interface for companies to use the forecasting tool.