daniel

nity

[Litz Fraaz @Fraazh]

Joined @ExileArms

[Litz Fraaz @Fraazh]

LITZ FAMILIA ❤️

[ESIV Analytics® @ESIV_analytics]

¡Gran noticia! 🎉 ESIV Analytics ahora es una marca registrada.

Gracias a nuestros clientes y aliados por confiar en nosotros. Seguimos comprometidos con transformar datos en decisiones estratégicas. 🚀💪

#ESIVAnalytics #MarcaRegistrada #AnálisisDeDatos #Consultoría

[MathMatize Memes @MathMatize]

Mathematicians seeing data scientists make millions of dollars by using basic applications of their theory

[ₕₐₘₚₜₒₙ @hamptonism]

How Many R’s are in the Word Strawberry?

[Pz MJ @quadhaed]

REJOINED @VirtueVE

[SoaR Zelta @SoaRZelta]

gotta love tiktok

[Probability and Statistics @probnstat]

Markov Chain Monte Carlo: Stochastic Simulation for Bayesian Inference by Dani Gamerman and Hedibert F. Lopes

Link to my upcoming Prob-Stat-ML Course: https://t.co/LcIoM7chuP

DM/Comment if you are interested!

[Joachim Schork @JoachimSchork]

Unlocking the full potential of your linear regression analysis starts with validating its suitability. Here’s a summary of how to validate assumptions before applying this model:

1️⃣ Linearity: The relationship between predictors (independent variables) and the response (dependent variable) should be linear. Think of a car traveling at a constant speed: the distance traveled increases proportionally with time.

2️⃣ Independence: Residuals (errors) must be independent of each other. Picture students taking an exam in separate rooms without communication. Each student's score is independent of the others.

3️⃣ Homoscedasticity vs. Heteroscedasticity: Residuals should have a constant spread across all levels of the independent variable. If the variability changes, like scores varying differently for easy versus hard exam questions, it indicates heteroscedasticity, which is problematic.

4️⃣ Normality of Residuals (Multivariate Normality): Residuals should follow a normal distribution, similar to the bell curve seen when rolling a fair die many times.

5️⃣ No Multicollinearity: Independent variables should not be highly correlated with each other. Imagine assessing the impact of both age and experience on salary: if they move together, it’s hard to distinguish their individual effects.

6️⃣ Measurement Error: Measurement errors can lead to unreliable predictions. It’s like weighing fruits on a scale that occasionally gives wrong readings.

The visualization originates from a recent post by @intelligentle__

#statistics #regression #statisticalmodeling #datascience

[Fermat's Library @fermatslibrary]

Roberto Carlos' free kick against France in 1997 is an extreme example of the Magnus Effect! He had to kick the ball with an angular speed of 14 revolutions per second (faster than some DVD players).

[Frank Nielsen @FrnkNlsn]

Some books on high-dimensional probability and statistics, including high-dimensional covariance matrices.

[Overleaf @overleaf]

Exciting news: Overleaf's PDF viewer now includes advanced navigation and zoom options. Experience a smoother, more efficient workflow! 🚀

[Mathieu @miniapeur]

How to be a PhD student:
1. Open 30 arXiv tabs
2. Never read any of them
3. Restart your browser
4. Repeat

[The PhD Place @ThePhDPlace]

Finishing your masters to start a PhD

[Rosana Ferrero 📈📊🙌 @RosanaFerrero]

🤣 Cuando un modelo de regresión lineal supera a un modelo de aprendizaje profundo
🧠 Especialmente cuando se buscan modelos simples, interpretables y rápidos de implementar.
#RStats #python #code #ML #DL #AI #IA #datascience #stats #analytics #deeplearning
RT @data_question

[Frank Nielsen @FrnkNlsn]

🎉A very unique textbook
"Information theory, inference and learning algorithms" by late Sir MacKay combining Information Theory with Machine Learning. Nicely written.

Chapter 27 on Laplace's method is only 2 great pages!

Book 🆓PDF available at:
👉https://t.co/wc62Wli1y3

[The Struggling Scientists @TheStrugglingS4]

But Peter didn't listen 🥲

[daniel @xNity_]

@klytage