Inter-arrival Time, Clustering, and Patterns
This article delves into the human tendency to seek patterns amidst randomness, exploring the interplay between signal and noise through my reflections.
Binary Outcomes, Risk Aversion, and Sorites Paradox (Part II)
This article explores the relationship between Binomial and Poisson distributions, using the Sorites paradox to highlight when Binomial approximations are “good enough” for real-world scenarios.
Binary Outcomes, Risk Aversion, and Sorites Paradox (Part I)
This article explores how increasing probabilities of success in a fixed number of trials can lead to negatively skewed distributions. It highlights the paradox that higher success rates may increase risk aversion due to greater sensitivity to potential losses.
Narratives, Chekhov’s Gun, and Agency
The article delves into the nature of cinematic heroism and character transformation, using the concept of normal distributions and weighting functions to model how characters' choices and paths evolve over time.
Searle’s Chinese Room, Reductionism, and Gaussian Mixture Models
The article explores John Searle's Chinese Room argument, delving into philosophical concepts of reductionism and emergentism. It applies these ideas to financial markets using Gaussian Mixture Models (GMMs) to analyse complex data, demonstrating the interplay between reductionist and emergent perspectives.
Expectations, Tails, and Poisson-Gamma
The article explores updating assumptions in dynamic environments using Poisson-Gamma models, emphasising the need for flexibility in variable settings and stability in predictable ones for effective decision-making.
Continuity, Jumps, and Measuring Progress
This article explores how progress is driven by both continuous improvement and significant "jumps." While continuous progress can become monotonous, jumps provide key milestones and clear markers of time and advancement.
Gradient Ascent, Global Maximum, and Complexity
This article explores strategies for progressing from point A to point B using the gradient ascent algorithm, comparing simple and complex game scenarios to highlight the challenges of reaching higher levels of achievement.