# Memory Augmented Deep Generative models for Forecasting the Next Shot Location in Tennis

## TL;DR

This paper introduces a memory-augmented deep generative model for predicting tennis shot locations and types, leveraging neuroscience-inspired memory modules to capture player behavior and adapt to match context.

## Contribution

It proposes a semi-supervised GAN framework incorporating episodic and semantic memory modules for modeling player behavior in tennis.

## Key findings

- Effective prediction of shot location and type
- Ability to analyze player style adaptation
- Demonstrated on 2012 Australian Tennis Open data

## Abstract

This paper presents a novel framework for predicting shot location and type in tennis. Inspired by recent neuroscience discoveries we incorporate neural memory modules to model the episodic and semantic memory components of a tennis player. We propose a Semi Supervised Generative Adversarial Network architecture that couples these memory models with the automatic feature learning power of deep neural networks and demonstrate methodologies for learning player level behavioural patterns with the proposed framework. We evaluate the effectiveness of the proposed model on tennis tracking data from the 2012 Australian Tennis open and exhibit applications of the proposed method in discovering how players adapt their style depending on the match context.