Machine-Learning on Harshvardhan /tags/machine-learning/ Recent content in Machine-Learning on Harshvardhan Harshvardhan /img/sharing-image.png /img/sharing-image.png Hugo en Sat, 18 Oct 2025 00:00:00 +0000 Enterprise-Scale Machine Learning for Demand Forecasting /foresight-paper/ Sat, 18 Oct 2025 00:00:00 +0000 /foresight-paper/ Our HP Inc. forecasting framework was featured in <em>Foresight: The International Journal of Applied Forecasting</em> (Issue 79, 2025) and recognized as a finalist at the International Institute of Forecasting’s Foresight Conference. <a href="https://www.harsh17.in/docs/papers/HP_Foresight_Paper.pdf">🔗 PDF</a> From Data to Decisions: Enterprise Demand Forecasting with Machine Learning /dissertation/ Sat, 31 May 2025 00:00:00 +0000 /dissertation/ My Ph.D. dissertation (University of Tennessee, 2025) develops a machine-learning-driven demand forecasting framework implemented at HP Inc., improving forecast accuracy by 34% and reducing inventory by 28%. <a href="https://www.harsh17.in/docs/2025_04_10_Doctoral_Dissertation.pdf">🔗 PDF</a> Artificial Intelligence and Data Sciences in Real-world Business /data-sciences-in-real-world/ Thu, 07 Sep 2023 13:00:00 +0000 /data-sciences-in-real-world/ A quick rundown of my doctoral research presented at the BA Forum 2023 Reflections from KDD 2023 /reflections-from-kdd-2023/ Mon, 14 Aug 2023 00:00:00 +0000 /reflections-from-kdd-2023/ My notes on talks I attended (mostly on LLMs) at 29th ACM SIGKDD 2023 at Long Beach, CA End-to-End Inventory Prediction and Contract Allocation for Guaranteed Delivery Advertising /kdd2023talk/ Mon, 07 Aug 2023 16:00:00 +0000 /kdd2023talk/ We proposed a novel end-to-end approach, the Neural Lagrangian Selling (NLS) model, to improve Guaranteed Delivery (GD) advertising by concurrently predicting ad impression inventory and optimizing contract allocation End-to-End Inventory Prediction and Contract Allocation for Guaranteed Delivery Advertising /kdd2023/ Wed, 07 Jun 2023 00:00:00 +0000 /kdd2023/ We proposed a novel end-to-end approach, the Neural Lagrangian Selling (NLS) model, to improve Guaranteed Delivery (GD) advertising by concurrently predicting ad impression inventory and optimizing contract allocation. The model incorporates a differentiable Lagrangian layer and a graph convolutional neural network to enable direct optimization of allocation regret and effective handling of various allocation targets and constraints. <a href="https://www.harsh17.in/docs/kdd2023/E2E_Paper.pdf">🔗 PDF</a> Joint Probability-based Dissimilarity Measure for Discrete Variables /joint-probability-based-dissimilarity-measure-for-discrete-variables/ Tue, 10 May 2022 00:00:00 +0000 /joint-probability-based-dissimilarity-measure-for-discrete-variables/ A Simple Method to Calculate Distance between Discrete Variable Basics of Text Mining in R /basics-of-text-mining-in-r/ Thu, 27 Jan 2022 00:00:00 +0000 /basics-of-text-mining-in-r/ Thinking of Text as List of Words Supervised Learning Using Baysian Decision Rule /supervised-learning-using-baysian-decision-rule/ Tue, 07 Sep 2021 00:00:00 +0000 /supervised-learning-using-baysian-decision-rule/ Python Functions for Bayesian Learning (COSC 522 Project)