Implementing Hardware Decision Tree Prediction: A Scalable Approach

摘要: Performance of data classification systems is one the most important aspect when involved volume, combined with classical computing approaches, does not match tight constraints on latency and throughput. Indeed, even though accuracy modern machine learning tools very suitable for adoption in many applications, they require computational resources elaboration time. In literature, a huge effort has been done to define new architectures several hardware implementations have introduced. this paper, we show implementation system based Decision Tree formally give demonstration its scalability terms required resources. At end, significant amount experimental evidences, prove that occupied area power consumption linear behavior against parameters.