Cluster Analysis of Merger and Acquisition Patterns in the Electronic Design Automation Industry Using Machine Learning Techniques

Abstract

Mergers and Acquisitions (M&A) are key drivers of the evolution of the Electronic Design Automation (EDA) industry through technology integration and market expansion. Here, we apply an AI-based clustering approach to unveil and explain the different waves of M&As. Two clustering algorithms are used: K-Means algorithm with the help of ML.NET to cluster the data based on some quantitative characteristics; and Density-Based Spatial Clustering of Applications with Noise (DBSCAN) that uses a density-based approach to detect clusters and outliers. Temporal transformation is used for the historical data and feature vectors are encoded using temporal metrics and a combined business model encoding. The approach that is proposed offers actionable insights by building feature sets from the number of days since the earliest event, the operational lifetime of the acquired companies, and a technological impact score or encoded business model data. These patterns are useful for industry stakeholders to understand the market consolidation trends without the need to understand the details of the underlying infrastructure.