For the python code, I used the Iris dataset which is available within the Scikit-learn package. It is a very small dataset (150 rows only) with a multiclass classification problem. As we are mostly focussing on hyperparameter tuning, I have not performed the EDA(exploratory data analysis) or feature engineering part and directly jumped into the model-building. I used the XGBoostClssifier algorithm for the model-building to classify the target variables. GridSearchCV is a function that comes in Scikit-learn's(or SKlearn) model_selection package.To use the GridSearchCV function, first, we define a dictionary in which we mention a particular hyperparameter along with the values it can take.

The entire world is transforming digitally and our relationship with technology has grown exponentially over the last few years. We have grown closer to technology, and it has made our life a lot easier by saving time and effort. Today everything is accessible with smartphones -- from groceries to cooked food and from medicines to doctors. In this hackathon, we provide you with data that is a by-product as well as a thriving proof of this growing relationship. Click here to head to the hackathon.

We are excited to announce our next MIIA event that will be hosted by PwC in V&A Waterfront, Cape Town on the 16th of August 2017 at 6pm. The theme of the event is "Data Science Applications to help businesses thrive in the Smart Technology Era" and we have speakers from PwC, Dimago, Cortex Logic and UCT. The Machine Intelligence Institute of Africa (MIIA) has also recently partnered with the Africa Data Forum and will be participating in the upcoming All Things Data Conference at the Westin, Cape Town on 29-30 August 2017. MIIA has decided to join forces with the Africa Data Science Association and will also have Board representation in this non profit organization that aims to ensure alignment between the academia and corporates on data science curriculum. We are excited to have many of the MIIA community members also participating in the upcoming Deep Learning Indaba at Wits, Johannesburg from 10-15 September 2017.