ITM UNIVERSE Student Patents

 

        KUNAL  ABHAYSHAH 

        DEPARTMENT: B.E CSE SEM-3, ITM UNIVERSE 

       TWO SINGLE INVENTOR INDIAN PATENTS  IN 2019

 

 

 1) Title : Analysis of efficiency of finite state machine in the light of shortest path and fuzzy rule

                Patent Number : 201911011077

                Filing Date :22/03/2019

Abstract : The finite state machine can be realised in the light of Zipf’s law and Hebbian Theory of learning. We claim that on the basis of fuzzy rule based optimized activation function, we can achieve time efficiency.

2) Title : A novel approach to realize reinforcement learning using finite state machine

                Patent Number : 201911011078

                Filing Date :  22/03/2019

Abstract : By virtue of deep-learning and association rule, a finite state machine can generate optimum throughput based on data processed by reinforcement learning.We further claim that fault tolerance in reinforcement learning plays a pivotal role  whereby there is no error propagation.

 

           STUDENT NAMAE: SIDDHANT BANE

           DEPARTMENT:   B.E CSE SEM-3, ITM UNIVERSE 

          TWO SINGLE INVENTOR INDIAN PATENTS  IN 2019

 

 

1) TitlePrediction of Campaign Donations on the basis of Compound Poisson Process Model

              Patent Number :  201911011074

              Filing Date :   22/03/2019

Abstract : In the process of estimation of campaign donations, there is a need to know when to expect the next donation and what  amount of donation is to be expected. In such a situation, where the arrival of the random variables and the size of the jumps is random, the Compound Poisson Process (CPP) has to be used. The CPP is a stochastic process that yields a random jump at a random interval from the Poisson distribution. In this  perspective the proposed inventor claims that CPP model plays a pivotal role in the process of predicting campaign donations. In addition to it, the inventor further claims that estimation of campaign donations can be governed by Poisson rate. (α)

2) Title :  Compound Poisson process (CPP) for product revalidation and its performance enhancement using Bootstrap aggregation.

                   Patent Number :  201911011075

                   Filing Date :   22/03/2019

Abstract : The Compound Poisson process is a stochastic process that yields a random jump at a random interval from the Poisson distribution. In the product  revalidation stage, design and redesign of a product is carried out using the feedback from focus groups and market trends. CPP model for product revalidation is cost effective and easily upgradable. Performance of CPP model can be improved using bootstrap aggregation.