pareto

CK automation actions to let users implement portable, customizable and reusable program workflows for reproducible, collaborative and multi-objective benchmarking, optimization and SW/HW co-design:

Multiobjective active learning with tunable accuracy/efficiency tradeoff and clear stopping criterion.

Tier-1 strategy-consultant frameworks (MECE, Issue Trees, Hypothesis-Driven, Pareto, So What?) packaged as Claude Skills + LLM-agnostic prompts. Drop in any LLM and get whiteboard-style structured analysis. Adapted from Analyst Academy on YouTube.

Implementation of NSGA-II in Python

A unified Python library for multi-objective feature selection algorithms, built for both research reproducibility and real-world applications.

A Python-based, open source, desktop application to import, read and analyse excel files in order to create and display multiple charts.

Multiple-criteria decision-making (MCDM) with Electre, Promethee, Weighted Sum and Pareto

Advanced choice modeling with multidimensional utility representations.

Python bindings for OptFrame C++ Functional Core

Source code for the paper "Toward Finding Strong Pareto Optimal Policies in Multi-Agent Reinforcement Learning"

WSQ Certified Lean Six Sigma Green Belt (CLSSGB) Training (TGS-2025055775) — 4-day / 32-hour DMAIC course with 25 hands-on activities, Excel datasets, slide deck, Learner Guide and Lesson Plan.

Multi-Objective Multi-Agent RL with non-linear utility functions

Discrete-tick simulator for KV-cache memory compaction policies in LLM inference servers. Compares NoCompaction, GreedyCompaction and ThresholdCompaction via 2D parameter sweep, Pareto frontier analysis and latency impact (P95/P99). C++20 + Python.

Pareto (Type I) distribution variance.

Pareto (Type I) distribution expected value.

Pareto (Type I) distribution standard deviation.

Pareto (Type I) distribution excess kurtosis.

(Completed) Machine Learning and Multi-Objective Evolutionary Algorithms to Solve Real World Engineering Problems (MultiObjectiveOptimisation and ML)

Pareto (Type I) distribution skewness.

No single LLM dominates every workload. LLM-OpsBench benchmarks GPT-5.4, Claude Sonnet 4.6, Gemini 2.5 Flash, and Llama 3.1 across four tasks, logging cost, latency, and quality per request. Includes Pareto analysis, ablation sweeps, and a Streamlit dashboard for actionable deployment decisions.