Name: SHAP Values & Waterpumps Competition
Start: 2025-09-25T18:00:00+12:00
End: 2025-09-25T20:00:00+12:00
Location: Grid AKL - Tech Cafe

This session will focus on applying machine learning to tabular data, using the Waterpumps competition as our practical example: [https://www.drivendata.org/competitions/7/pump-it-up-data-mining-the-water-table/](https://www.drivendata.org/competitions/7/pump-it-up-data-mining-the-water-table/). If you've been wanting to learn about a new Machine Learning technique, this is your chance to get hands-on experience.

We will also be discussing SHAP (Shapley Additive Explanations), a powerful method for interpreting the predictions of machine learning models by explaining how individual features contribute to the model's output.

After building a model to predict the status of a water pump, how do you explain its decisions? SHAP values provide the answer. In this session, we'll explore how you can use SHAP to understand *why* your model makes a specific prediction for a given water pump. For example, you could determine whether features like `construction_year`, `gps_height`, or `water_quality` were the most influential factors in classifying a pump as "functional" or "non-functional." This provides deep, practical insights beyond simple accuracy scores.

Shap Resources

* **SHAP Chapter** from the highly regarded ML Interpretability book - [https://christophm.github.io/interpretable-ml-book/shap.html](https://christophm.github.io/interpretable-ml-book/shap.html)
* **Explain Your Model with the SHAP Values** article - [https://towardsdatascience.com/using-shap-values-to-explain-how-your-machine-learning-model-works-732b3f40e137](https://towardsdatascience.com/using-shap-values-to-explain-how-your-machine-learning-model-works-732b3f40e137)
* **SHAP video** - [https://www.youtube.com/watch?v=VB9uV-x0gtg](https://www.youtube.com/watch?v=VB9uV-x0gtg)
* **SHAP GitHub Repository** (and official documentation) - [https://github.com/shap/shap](https://github.com/shap/shap)

Please try to review some of the materials before attending. This will facilitate a more productive discussion as participants will have a shared foundational understanding.

Who is this meeting for?
This meetup is for individuals interested in Data Science and Machine Learning, particularly those involved in or curious about tabular data and model interpretability.

Discussion Format:
We will follow a reading group format, providing an opportunity for attendees to share their insights and ask questions. Participants are welcome to pass if they prefer to listen.

Meeting Schedule:

* 6:00 PM - 6:30 PM: Arrive / Informal Networking
* 6:30 PM - 6:35 PM: Introductions
* 6:35 PM - 7:25 PM: Discussion
* 7:25 PM - 7:30 PM: Nominate and vote for the next meeting's topic
* 7:30 PM: Meeting Concludes

We look forward to seeing you there.

Data Science Discussion Auckland

Technology

Data Mining

Big Data

Machine Learning

Data Analytics

Predictive Analytics

Computer Programming

Data Science

SHAP Values & Waterpumps Competition

Grid AKL - Tech Cafe

Share

Data Science Discussion Auckland

SHAP Values & Waterpumps Competition

Data Science Discussion Auckland

Details

Members are also interested in