Being a researcher demands commitment, sustained effort, and a high level of inspiration. Ms Lakshmi Bhargavi from 3rd-year Computer Science Engineering has presented a paper titled “Application of distributed back propagation neural network for dynamic real-time bidding” at the 11th International Advanced Computing Conference (IACC-2021), hosted by the University of Malta, Europe. It is a reputed conference indexed in Scopus and DBLP, having H-index 25.

The research is based on the backend of ad placement on websites which involves finances. The process involves finding the best deal between the dealer and the supplier. In the present system, the bid is prefixed, thereby reducing the possibility of optimal budget utilisation. In comparison, Ms Lakshmi’s research uses an ML algorithm, which is dynamic and learns from the previous bids. This research has resulted in 15% lower costs for the suppliers, thereby saving a lot of money and resulting in a better system.

Abstract — Programmatic buying, popularly known as real-time bidding (RTB), is a key ascendancy in online advertising. While data has become essential for targeting and ad performance, data businesses have become difficult to differentiate due to their proliferation, as well as limitations of attribution. This provides an opportunity for Big Data practitioners to leverage this data and use machine learning to improve efficiency and make more profits. In such an opportunity, the research came up with an application of a machine learning algorithm, distributed back propagation neural network, d-bpnn, to predict bid prices in a real-time bidding system. This paper depicts how d-bpnn is used to achieve less eCPM for advertisers while preserving win rate and budget utilisation.

The 11th International Advanced Computing Conference (IACC-2021) was hosted by the University of Malta, Europe, with an H-index of 25. The conference is indexed in Scopus and DBLP and in collaboration with Springer. A few selected papers will be published in SCOPUS/SCI Indexed journals. The presentation was held on 19th December 2021. It was an ideal platform for people to share views and experiences in Futuristic Research Techniques.

Let’s hear from Ms Lakshmi:

My university has been with me in every step taken towards this conference. I would like to thank the mentoring of Dr Priyanka throughout the writing and presentation of the paper. The immense support of SRM AP management, my professors, HOD, Pro VC sir and VC sir made me reach the level to write a paper confidently and show my knowledge to the world.
I feel honoured to present a good paper at a global conference. The experience and connections I made through this conference are priceless. It gave me new insights into several other technical domains. I believe I gave my best at the unique opportunity given to me and hopefully will continue to deliver good work in future too.

SERB-DST projects aim to build up the best systems that would match the best global practices in the area of promotion and funding of basic research. Dr Murali Krishna Enduri, Assistant Professor, Department of Computer Science Engineering at SRM University-AP is yet another faculty member who has obtained a project with a total outlay of ₹18 lacs for a duration of three years. The project is sanctioned under the scheme of Teachers Associateship for Research Excellence (TARE) of SERB-DST, Government of India.

In the duality theory, the dual problem is the problem of checking the duality of a pair of monotone Boolean expressions in disjunctive normal form. Problem: DUAL Input: The complete DNF of two monotone Boolean functions, f and g. Output: If f is dual of g. Whether the problem DUAL admits a polynomial-time algorithm has been one of the challenging open problems in the field of Duality theory of Boolean function for the last 35 years. It is one of the few problems whose polynomial-time solvability is still unknown. So, this problem is important in complexity theory due to its unknown complexity status and it plays a central role in various applications arising in computational logic, data mining, reliability theory, artificial intelligence and game theory etc. The project goal is to solve the dual problem for an interesting class of Boolean functions. Improving the existing complexity results of the DUAL problem for a particular class of Boolean functions is a challenging task.

Few applications of the project are as follows:
Type error diagnosis: Type error diagnosis is the task of generating an explanation for some error. It requires finding all minimal unsatisfiable subsets of a given set of constraints (representing the error) which can be managed via solving the computational variant of Dual in its minimal transversal formulation.

Computational medicine: Optimal vaccination strategies are given a subset of initially infected individuals from a population of individuals and assumptions about disease transmission. The task of computing inclusion minimal vaccination strategies can be solved using the computational variant of Dual in its transversal hypergraph formulation.

The project will be carried out in collaboration with IIT Madras (Dr Jayalal Sarma, Associate Professor, Department of Computer Science & Engineering, Indian Institute of Technology Madras, Chennai, India.)