Pathway’s Post

Exciting news from ICSE 2024! Our team is thrilled to be presenting our LLM solution in BWX at this year's International Conference on Software Engineering. Even better, our paper was selected to be presented at CAIN, the AI conference within ICSE. We can't wait to share our case with fellow industry professionals. Stay tuned for updates! #ICSE2024 #CAIN #AI #SoftwareEngineering

🌟 Meet our hackathon winners pt.4 🌟 What advice do they have for other data scientists and what motivated them to dive into responsible data science? We interviewed them to find out. Adarsh is pursuing B. Tech from the Indian Institute of Technology, Guwahati and is fascinated by the various domains of data science and the application of AI for societal betterment. He’s won the notebook prize at the Harvard OpenDP hackathon where he approached the problem from clever angles, broke the project into logical easy-to-follow sections, and showcased his creativity and statistical intuition. 🙌🏽 🔍 Motivation “The Harvard Open DP Challenge seemed like an amazing opportunity to learn a new key skill of Differential Privacy, which, obviously, has many real world applications. This was a great opportunity to enhance my skills and explore a new domain.” 🤓 Approach “I focused my approach around trying various methods to extract information from seemingly uninformative and noisy fields. Trying various arbitrary things like, say, taking the first 6 digits of the phone number, which made the previously noisy and useless phone number field just a bit more useful.” 🔑 Advice “As an up-and-coming data scientist myself, all I can say is just put yourself out there, just try to learn as much as you can, and participate in as many things as you can. Embrace this journey of continuous learning and trust the process.” You can check out his solution here: https://lnkd.in/dUVEkUZb Want to get started with Eyes Off Data Science? Sign up to our platform: https://lnkd.in/deQ4XriY #DataScience #Privacy #Hackathon #EyesOffMyData

Happy to announce the publication of my paper titled "Generative vs Intent-based Chatbot for Judicial Advice" in the conference proceedings of IEEE Xplore. The work was presented at the 2nd IEEE International Conference on Interdisciplinary Approaches in Technology and Management for Social Innovation (IATMSI-2024). The paper presents a generative chatbot and an Intent-based chatbot aimed at providing judicial advice to Indians and explore and compare the two approaches on the basis of various factors such as nature of responses, response quality, handling changing scenarios, training and data requirements and user experience in the context of offering judicial advice specific to Indian laws. Thanking my guide, Medha Wyawahare and my teammate Sarang Zanwar for the shared efforts in making this possible. Link to access the paper: https://lnkd.in/dFJ4-6cb

#YDC announces the General Availability of YDCGPT and YDC #AIEconomics Center for GenAI value quantification, see the full release here Michael C. https://lnkd.in/ey2swaTz

Great presentation of MFDM and HyperOps from TCS Research .

Finally completed a course from UPGRAD : Artificial Intelligence in the Real World

Through Knowledge Sharing to Singularity, Accelerated By LLMs https://bit.ly/3Tjnd1T

In the latest update (version 1.2) of my app (Nabu), I've introduced some exciting new features: I incorporated additional tests into the regression analysis. These enhancements enable users to identify which observations most significantly affect the model's coefficients using Cook's distance, and to assess how extreme certain values are in relation to the mean, known as leverage. I also introduced new visuals to complement these new enhancements. Cook's distance is a measure used in regression analysis to identify influential observations. These are data points that, if removed, would significantly change the estimated coefficients of the model. A high Cook's distance indicates an observation that is a potential outlier or has high leverage. Leverage is how "unusual" or "extreme" an observation's predictor (independent variable) values are compared to the rest of the data. In the upcoming update, I plan to integrate both supervised and unsupervised machine learning models into the application, complete with all the necessary tests. #dashboarddesign #dashboard #dataanalytics #dataanalysis #machinelearning #statistics #analysis #analytics #ai #aicommunity https://lnkd.in/eqJFxAps

How did the ECG generation project turn out? 🔬 Unfortunately, the results generated by a GAN model were insufficient. This failure taught me a lot, starting with being critical with online papers and being efficient in organising the development process, especially for a model as far from easy to train as the GAN. I want to share the presentation of the project, in which my Tobia Bacchiddu,@Edoardo Luziatelli teammates and I explored the general topic of Data Augmentation, to which I will devote myself fully in a month's time. Special thanks go to Professor @Hugues Bersini and the universities Université libre de Bruxelles,Vrije Universiteit Brussel who gave me the opportunity to listen to several speakers about ‘Current trends in AI’. Among the many interesting topics, I would like to emphasise the importance of future research for RAG, MLOps, LLMOps, FSM and Hierarchical Alphabet Automata, RegTech. 📔 Github repository: https://lnkd.in/d3H6dzRa #DataAugmentation #FutureResearch #TrendsinAI #GAN #DataGeneration

The excellent State of ZK report by Zero Knowledge Validator is publicly available 👀 Our favorite part: "ZK Score by Omer Shlomovits and Lucy Williams is a proposed metric system for evaluating Zero Knowledge Proof (ZKP) technologies, focusing on efficiency at the hardware level. It measures the number of proofs generated per joule of energy consumed, providing a standardized way to compare #ZK tech stacks. The goal of ZK Score is to offer a tangible, hardware-centric benchmark, akin to those in #AI technologies, to assess the performance of various #ZKP implementations. By measuring proofs-per-Joule, ZK Score aims to quantify the energy efficiency of ZKP systems, enabling a fair and objective comparison." Link to ZK Score blog post in the 1st comment 👇 Link to Zero Knowledge Validator's State of ZK Report in 2nd comment 👇