FIFA Position Predictor - ML Hackathon [Predictive Exp.]

This experiment infers the optimal position on the football pitch for a player given their footballing attributes. The dataset used was obtained from the renowned game FIFA's player database. **DATA** The database consists of the attributes of professional footballers and the target attribute is their respective position. The dataset contains 22 features and 1 label column. The features are all integer values on the scale of 1 to 100, 100 being the highest. The following diagram shows an excerpt from the dataset: ![nn][1] **EXPERIMENT** The model used here is Multiclass Decision Jungle. The dataset is initially processed by an R script which refines it. The data is then split in the ratio 0.7 to 0.3 into training and testing examples. The training data is fed into the classification model, which is then scored using the testing data. This scored classification is finally evaluated with standard metrics. Decision jungles are a recent extension to decision forests. A decision jungle consists of an ensemble of decision directed acyclic graphs (DAGs).By allowing tree branches to merge, a decision DAG typically has a lower memory footprint and a better generalization performance than a decision tree, albeit at the cost of a somewhat higher training time. Decision jThey perform integrated feature selection and classification and are resilient in the presence of noisy features. [1]: https://drive.google.com/file/d/1GwXs1WH4CKUB02Yt9trI7OFF0GI4alq1/view?usp=sharing [2]: