Winston Li’s Post

I am speaking at the 8th Annual Toronto Machine Learning Summit (#TMLS) 2024. The title of my talk is Dynamic Huff's Gravity Model with Covariates. In this presentation, I will discuss an innovative #retailanalytics technique based on the Gravity model for predicting and understanding how likely are customers going to visit certain retail locations and what are the main factors that drive them to visit these locations. My talk will be happening on Mon Jul 15, 2024 12:05pm – 12:35pm (EDT). https://lnkd.in/gdqTEVRj Please let me know if you'd like a discount code for purchasing a ticket! #tmls2024 #machinelearning #datascience #analytics #predictiveanalytics #syntheticsociety

📣 Exciting News Alert! Check out our latest blog post on "CCNETS: A Novel Brain-Inspired Approach for Enhanced Pattern Recognition in Imbalanced Datasets." This study introduces CCNETS, a cutting-edge generative model-based classifier designed to tackle the challenge of imbalanced datasets in pattern recognition. With its unique brain-like information processing, CCNETS has demonstrated superior classification ability, outperforming traditional models in handling imbalances and enhancing data generation. Dive into the details and explore the potential of CCNETS in revolutionizing machine learning. Read the full article at: https://bit.ly/3vqvjfH [cs.LG] #PatternRecognition #MachineLearning #CCNETS #DataGeneration

If you're heading to Weights & Biases #FullyConnected2024 Conference this week, make sure to catch Seth Levine's session on the benefits of different approaches to AI and how we're applying them to customer experience conversations

Looking forward to presenting at Weights & Biases Fully Connected Conference this Thursday! Join me for my talk: “Human x Machine: From Models to Products” where I share my perspective on the state of machine learning in industry and the Loris story 🎉 #FullyConnected2024

InfoHarmony ML: Unleashing Predictive Precision through Mutual Information and Linear Regression 📈 Embark on a journey with InfoHarmony ML, where the symphony of Mutual Information Classification, Mutual Information Regression, and Linear Regression converge to orchestrate predictive precision. Unravel insights, predict outcomes, and navigate the data landscape with this innovative fusion of algorithms. Explore the predictive power on GitHub: [https://lnkd.in/dnvWQJ4g] #MachineLearning #InfoHarmonyML #PredictiveAnalytics 🚀🔍

Decision-ready data 🛰️🌎📊 Check out the Information Technology and Interoperability (IT&I) call with Brian Goldin from Voyager Search. As organizations grapple with information overload, timely and reliable insights remain elusive, particularly in disaster scenarios. Voyager's participation in the The Open Geospatial Consortium (OGC) Disaster Pilot 2023 aimed to address these challenges by streamlining data integration and discovery processes. Leveraging innovative #EarthScienceData conditioning and enrichment techniques, alongside #ML models, Voyager transformed raw data into actionable intelligence. Through operational pipelines, Brian and his team linked diverse datasets with machine learning models, automating the generation of new observations to provide decision-makers with timely insights during critical moments. ➡️ https://lnkd.in/g_4zm-z8

Explore the innovation of Bayesian Flow Networks (BFNs) in generative modeling. BFNs redefine data relationships and optimize generative processes. https://lnkd.in/gUkApeCm #GenerativeModeling #BayesianFlowNetworks #Innovation

🌟 Day 31 of 51-Days Data Structure Learning Challenge by CodeRescuers! 🌟 Today, immerse yourself in the brilliance of Radix Sort, a groundbreaking algorithm that sorts numbers by their individual digits, offering unparalleled efficiency in data management and analysis! 🚀 Explore its unique approach to sorting and unlock the secrets to mastering large datasets with precision! 💻 #CodeRescuers #DataStructures #RadixSort #Algorithm #CodingChallenge #TechTrends

Hey everyone! Guess what? We’re making magic happen! Things are getting super exciting with our project. Today, I focused on training several machine learning models, including Logistic Regression, Decision Trees, and Random Forests. This step is crucial as it converts our preprocessed data into predictive power. I evaluated the models using accuracy, precision, recall, and F1-score. The Random Forest model emerged as the top performer, demonstrating high accuracy in detecting network intrusions. Now, it's time to get this model deployed and working its magic in real-time network monitoring.  #Amdari21DaysDatachallenge #Datastar #DataScience #NetworkIntrusionDetection #Day4

So amazing how quickly the LLM landscape continues to evolve, with the release of Meta's Llama 3 open source model rivaling the best proprietary models out there. Now, anyone has access to the guts of one of the best models out there - the code, architecture, training process - for free, democratizing access to powerful language AI capabilities. This could catalyze a new wave of creative breakthroughs. I'll admit, I never expected to be this enthusiastic about a Meta product. But the democratizing potential of Llama 3 has me genuinely intrigued. #AIRevolution, #OpenSourceAI, #LanguageModels, #MetaAI, #Innovation, #TechForGood, #AIandEthics https://lnkd.in/eqkh7bTv