A pleasant summer salad

On weights and biases.

Using ILP for a decision making process.

No, the other inference!

Notifying me about relevant blind spots of my life.

If there's something we can learn from Snickers, it's that peanuts and chocolate work.

I started using Emacs' OrgMode to log my days.

After witnessing conventionally spectacular achievements, I embraced the arbitrariness of goals.

I explored OpenAI's gym library to model Sudoku as a Reinforcement Learning problem.

I compiled lists of media I have consumed in 2021.

I looked into my running volume of 2021.

I modeled solving a Sudoku board as a constraint optimization problem.

I modeled solving a Sudoku board as a graph problem.

I compiled lists of media I have consumed in 2020.

I looked into my running volume of 2020.

I contemplated on the usefulness of past observations when assessing normative visions.

I realized that I also care about speed aside from effectively being fast.

Don't buy beans for a hobby roaster.

I tried to model image denoising as a graph problem and implemented a naive optimization for it.

I built a Telegram chatbot which lets friends add music to my discovery playlist.

I built a Chrome extension which seems to work more or less.

I embarked onto the promising adventure of translating social complications into math.

Why am I defaulting to buying things I like to see?

Naming things remains hard.

I looked into my running volume of 2019.

I used Markov chains to generate images.

I tried to find a simple solution to a simple math problem.

I believe I could profit from being questioned more, with good intentions.

I compiled some improvements to my workflow.

I tried to illustrate the process of turning an already useful randomized algorithm into a conventional, non-random algorithm.

I tried to illustrate the usage of scikit-learn pipelines.

I tried to come up with a simple yet long-term system to log my running.

I contemplated on the dimension of specificity.

Is changing the currency of time a useful regulator?

I illustrated increasingly elaborate possible approaches for scenarios akin to the 'Netflix Prize'.