"I believe in intuition and inspiration. Imagination is more important than knowledge. For knowledge is limited, whereas imagination embraces the entire world, stimulating progress, giving birth to evolution. It is, strictly speaking, a real factor in scientific research.”
― Albert Einstein 1929, quoted by Dr. Boris Kovalerchuk in Visual Knowledge Discovery and Machine Learning 2018.

Double Bachelor's of Science senior student majoring in Computer Science and Applied Mathematics. Active research assistant at the Visual Knowledge Discovery and Imaging Lab of Central Washington University. Former system administrator, full-stack software developer, and web-developer. Next, graduate studies in Computational Science focusing on visual machine learning. With a goal of building trusted expert artificial intelligence systems.

Solving problems at the intersection of machine learning and data visualization by blending machine computation with human cognition. Computing with data visualizations directly, instead of just making visuals. While integrating visual approaches with conventional machine learning methods.

Oxford Dictionary, "Intelligence", noun, 1. the ability to acquire and apply knowledge and skills. Research goals: Aquire knowledge by visual knowledge discovery with multidimensional lossless visualizations, then apply aquired knowledge to build skills that solve difficult tasks more interpretably and accurately.

If you find my work of interest or benefit, then please consider a supporting gesture through , thank you.

In no preferential order.

Chronologically ordered from current at top to first at bottom.

3. "Synthetic Data Generation with Visual Knowledge Discovery"

• Authors: Alice Williams and Dr. Boris Kovalerchuk.
• Status: Actively researching at the CWU Visual Knowledge Discvery and Imaging Lab.
• Contributions: Expanding on synthetic data generation methods with interactive algorithms paired with Generative Adversarial Networks (GANS). Expanding software with relabeling and GAN data synthesization, in support of theorized algorithms. Exploring GAN deficiencies and seeking ways to integrate visual improvements.
• Topics: Synthetic Tabular Data Generation, Image to tabular data preprocessing, General Line Coordinates, Hyperblock Model Representation, Generative Adversarial Networks, Multi-Row Parallel Coordinates, Imagery Visualization
• Developed software: Dynamic_Coordinates_Vis_System, ML_Classifier_Comparison_Tool, DCGAN_Custom_Architecture_Builder_and_Image_Synthesizer

2. "Interpretable Computational Visual Knowledge Discovery and Boosting Classification Models with Human-in-the-Loop"

• Authors: Alice Williams & Dr. Boris Kovalerchuk.
• Status: Preparing camera ready version, accepted for presentation at AI Human Computer Interfaces 2025 in thematic area of Artificial Intelligence in HCI.
• Contributions: Visual solution to the 2003 Gödel Prize problem of AdaBoost classifier by visual observation of options for different classifiers before selecting them for boosting allowing alternative ways to build stronger classifiers, including directly building classifiers in a lossless visualization space and visualizing classifier uncertainty, applying visual knowledge discovery methods to extract data representational rules and improve classification models by rebuilding them with interpretable hyperblock representations.
• Topics: Human-Centered AI, Machine Learning, Multi-Class Classification, Feature Engineering, Explainable AI, AI Risk Mitigation, Human-in-the-Loop, Multi-Dimensional Distributions, General Line Coordinates, Concentric Coordinates.
• Developed software: Java_Tabular_Vis_Toolkit, HyperblockParser, InLineCoordinatesCoefficientSolver

1. "Synthetic Data Generation and Automated Multidimensional Data Labeling for AI/ML in General and Circular Coordinates"

• Authors: Alice Williams & Dr. Boris Kovalerchuk.
• Status: Published in IEEE proceedings and presented at IV2024 in track of Artificial Intelligence and Visual Knowledge Discovery, earned the 'Best Paper Award'.
• Contributions: Proposed an algorithm and implemented interactive software for labeled synthetic data generation using former General Line Coordinate methods and newly developed Circular Coordinates both Static and Dynamic with multi-class visualization and parameterized class discrimination. Addressed data balancing, demonstrated deficiencies of popular SMOTE (Synthetic Minority Oversampling Technique), and showed improvements to classifier performance across 14 standard machine learning classifiers.
• Topics: Synthetic Data Generation, Automated Data Labeling, General Line Coordinates, Circular Coordinates, Parallel Coordinates, Shifted Paired Coordinates, Tabular AI/ML Data, Multidimensional Data Visualization, Visual Knowledge Discovery.
• Developed software: Dynamic_Coordinates_Vis_System

I'm always open to discussing new projects or opportunities. Feel free to reach out or connect with me!