Applied Data Science

The unsung heroes 😭 #datascience #ai

🍼 Got my daughter a gift for her 1st epoch. Hoping her accuracy starts improving soon though. Does anyone have any tips for training this kind of model? #machinelearning

Modern problems require modern solutions 🔨 #datascience #llm #gpt #ai

The next 18 months are going to be wild.

After spending an entire day cleaning free text data for LLMs 🧼 #chatgpt #ai

@levikul09 Great breakdown! Thanks for sharing.

There are an infinite number of matrices, but some are special. Certain matrices have important characteristics and useful properties. Here are a few 🔽 1⃣ Square and Rectangle Square matrices have the same number of rows as columns. Rectangular matrices are tall if the number of rows > number of columns. And they are wide if # of rows < # of columns. 2⃣ Diagonal Diagonal matrices contain non-zero values only on the diagonal. Note: Diagonal elements can also be zero! np.diag() creates a diagonal matrix from a vector or list of values. 3⃣ Triangular A triangular matrix contains zeros either above or below the diagonal. Upper triangular matrices have zeros above, Lower triangular matrices have zeros below the diagonal. np.triu() and np.tril() create Triangular matrices from a given matrix. 4⃣ Identity The identity matrix is a square diagonal matrix where all diagonal elements are 1. It is indicated by Capital I and a number showing the size. np.eye() creates identity Matrices 5⃣ Zeros and Ones The zeros matrix is a matrix of all zeros. The ones matrix is a matrix of all ones. np.zeros() and np.ones() will create them. ___ That's it for today. I hope you've found this Tweet helpful. Like/Retweet for support and follow @levikul09 for more Data Science content. Thanks 😉

@impratik88 Cool breakdown! There is also some interesting research around quantifying serendipity in a recsys. eugeneyan.com/writing/serend…

Recently, while experimenting with a recommendation system project, I found myself using a variety of evaluation metrics. So I compiled a list of metrics that I found helpful. #MachineLearning @pratikaher88_98557/evaluation-metrics-for-recommendation-systems-an-overview-71290690ecba" target="_blank" rel="nofollow noopener">medium.com/@pratikaher88_…

I'm sure the user table had rows when I started work this morning... #datascience

@PythonMaps Great visualization!

Forests of South America! This map shows the percentage tree cover at a 30 × 30 meter resolution. #Python, #Matplotlib, #Numpy. #Data #DataScience #geography #SouthAmerica

@truepurpose_ 😂

python devs (mlops, data engineering, data science) all over the world relate

Metric of the Day EBITDA #infographic 😎

"Don't worry, it's simple." - A PM who spent too much time on LinkedIn this week. 🤦 #ai #startup