pantelis / artificial-intelligence

Hi there 👋

• 🔭 I’m currently working on Situational Awareness, Drones and Unmanned Surface Vehicles (USV) and Computer Vision applications.
• 🌱 I’m currently learning how to optimize local LLM deployments.
• 👯 I’m looking to collaborate on developing SaaS AI-enabled applications.
• 🤔 I’m looking for help with Web UI frameworks.
• 💬 Ask me about AI and ML
• 📫 How to reach me: Twitter: monogioudis, LinkedIn: pantelis
• ⚡ Fun fact: Every August I effectively disconnect for 2 weeks vacationing in the island of Chios, Greece.

Bio

I blog at aegean.ai and you can read my short bio here.

artificial-intelligence's Issues

Issue on page /aiml-common/lectures/optimization/maximum-likelihood/_index.html

Outdated OLS link. Replace with https://www.ccs.neu.edu/home/radivojac/classes/2021springcs6140/notescs6140linearregression.pdf

Issue on page /aiml-common/lectures/pgm/pgm-intro/_index.html

We hinted that assigning an object an a symbolic identity.

typo in Introduction to Probabilistic Reasoning.

Issue on page /aiml-common/lectures/pgm/recursive-state-estimation/_index.html

Issue on page /aiml-common/lectures/pgm/recursive-state-estimation/_index.html

What is PGM? The website has no explanation for it

Issue on page /aiml-common/lectures/ai-intro/course-introduction/_index.html

Final link titled "This" does not route. Updated link should be: https://blog.unity.com/engine-platform/training-a-performant-object-detection-ml-model-on-synthetic-data-using-unity

Issue on page /aiml-common/lectures/entropy/_index.html

I cannot understand what xt is from just the webpage

Issue on page /aiml-common/lectures/logical-reasoning/propositional-logic/_index.html

Stench when the Wumpus is in an adjacent square — directly, not diagonally

Stench when the Wumpus is in an adjacent square — adjacent, not diagonally

Issue on page /aiml-common/lectures/pgm/bayesian-inference/_index.html

Clarify that the Bayesian formula given here by David MacKay

"In this formula H is the overall hypothesis space, so there is no point in putting this as a p(H|xyz). I would read the formula as given in the lecture where a portion of the hypothesis space refers to the prior model of w (isotropic gaussian) and another to the linear relationship between x and y manifested via the conditional mean of p(y|x)=wo+w1*x."

Issue on page /aiml-common/lectures/ai-intro/course-introduction/_index.html

This is a demonstration of object detection using Omniverse - a photorealistic 3D creation and simulation environment.

is the line that has the dead link

Issue on page /aiml-common/lectures/nlp/rnn-language-models/_index.html

To train an LSTM language model

We start with big corpus of text which is a sequence of tokens
where T is the number of words / tokens in the corpus.

Every time step we feed one word at a time to the LSTM and compute the output probability distribution Missing argument for \mathbf, which is, by construction, a conditional probability distribution of every word in the dictionary given the words we have seen so far.

The loss function at time step is the classic cross entropy loss between the predicted probability distribution and the distribution that corresponds to the one-hot encoded true next token. $Missing argument for \mathbf$

Average all the t-step losses