AI Tech&Review site "Synced" examines "The Cake Analogy 2.0" and my take on self-supervised learning in my ISSCC keynote.
https://syncedreview.com/2019/02/22/yann-lecun-cake-analogy-2-0/

✴️ @AI_Python_EN

Image Translation with Tensorflow


#Pix2Pix is an Image-to-Image Translation with Conditional Adversarial Networks.

It can prove to be effective at synthesizing photos from label maps, reconstructing objects from edge maps, and colorizing images, among other tasks

Check out the original paper(https://lnkd.in/fFAm8YK) if you are interested in implementation detail, it shows more example usages for cGAN like "map to aerial", "day to night" et.

Here is a Tensorflow implementation of the same by Christopher Hesse(https://lnkd.in/f7ivy95)

#deeplearning

✴️ @AI_Python_EN

Based on book of "New Document.docx"
✴️ @AI_Python_EN

Thank you all for your support!
We're 1000 members now :)
✴️ @AI_Python_EN

Myth: Being a data scientist is applying machine learning 100% of the time

Fact: Being a data scientist is applying machine learning 5% of the time

The other 95% is spent:

- Understanding the Business Problem & Communicating with Domain Experts, 20%

- Working with data: Cleaning, Manipulating, Visualizing, Processing, Transforming, Understanding, 60%

- Communicating Results: Reporting, Slide Decks, and Apps, 15%

—-

Key Point - If you want to be a great data scientist, focus on where you will spend the most of your time. Communication, Business Understanding, Data Manipulation & Visualization

✴️ @AI_Python_EN

Uncovering and Mitigating Algorithmic Bias through Learned Latent Structure

Amini et al.: https://lnkd.in/e5Ybyfa

#artificialintelligence #deeplearning #machinelearning

✴️ @AI_Python_EN

💡 What is a p-value?

When testing an hypothesis, the p-value is the likelihood that we would observe results at least as extreme as our result due purely to random chance if the null hypothesis were true.

💡 What does it mean when a p-value is low?

When the p-value is low, it is relatively rare for the our results to be purely from random variations in observations.

Because of this, we may decide to reject the null hypothesis. If the p-value is below some pre-defined threshold, we say that the result is "statistically significant" and we reject the null hypothesis.

💡 What value is most often used to determine statistical significance?

A value of alpha = 0.05 is most often used as the threshold for statistical significance.

#datascience #statistics

✴️ @AI_Python_EN

Different Sorting Algorithms and how they work.

Source- Reddit

An awesome colllection of Deep Learning tutorials (on Github)-
https://lnkd.in/fyAFyK6

✴️ @AI_Python_EN

Top 10 NLP Concepts to analyze text data

Interested in NLP? Get familiar with these 10 algorithms before you get started:

1. tf-idf - https://lnkd.in/ghfqfm7
2. N-grams - https://lnkd.in/gCDChaT
3. Stemming - https://lnkd.in/gwHuE68
4. Lemmatisation - https://lnkd.in/gRU8Q5m
5. Cosine similarity - https://lnkd.in/gEMj9hp
6. Bag-of-words - https://lnkd.in/gzv7NDX
7. Word2vec - https://lnkd.in/gV2yEsn
8. LDA - https://lnkd.in/gF2qcnJ
9. Edit distance - https://lnkd.in/gy3wU5H
10. LSTM - https://lnkd.in/gu9H9vM

Get familiar with these concepts at a high level, and then grab a dataset and start playing around. You'll learn a lot more by implementing these concepts with real data than just reading and studying forever.

Here are two great sources to grab free text datasets:
👉 https://lnkd.in/gABJX4w
👉 https://lnkd.in/gFR9njn

Remember to start simple and then iteratively build and test from there, not every model required deep learning
👉 Get familiar with these 10 concepts and and you'll be ready to conquer challenging NLP problem.

#datascience #machinelearning #nlp #deeplearning #algorithms

✴️ @AI_Python_EN

#NLP is among the hottest and most interesting fields in #datascience. Check out these 5 in-depth and hands-on tutorials to learn #NLP:

1. The Essential NLP Guide to Solve Top 10 Common NLP Tasks - https://lnkd.in/fiXS5Rj
2. Practical Tutorial for Regular Expressions in #Python - https://lnkd.in/fXw-Rdz
3. A Gentle Introduction to #TopicModeling - https://lnkd.in/fDXmt4n
4. Comprehensive and Intuitive Guide to #WordEmbeddings - https://lnkd.in/fvRrFhA
5. #TextClassification using #ULMFiT and #fastai Library in Python - https://lnkd.in/f7bu8jM

And test your #NaturalLanguageProcessing knowledge on this challenging question set!
30 Questions to test a data scientist on Natural Language Processing - https://lnkd.in/fpWBZUh

✴️ @AI_Python_EN

Generating Music With Artificial Intelligence
http://bit.ly/2HnrmO6
#DataScience #MachineLearning #ArtificialIntelligence

✴️ @AI_Python_EN

RNN-Based Handwriting Recognition in Gboard

Blog by Sandro Feuz and Pedro Gonnet: https://lnkd.in/eUbtqyi

#artificialintelligence #deeplearning #machinelearning

✴️ @AI_Python_EN

Consolidation of statistical tests in single page

#statistics

✴️ @AI_Python_EN

Just for laughs - Difference between Machine Learning and Artificial Intelligence (how many of you all have felt this? :)

✴️ @AI_Python_EN

IBM Cognitive classes is one way to learn Data Science and Machine Learning, and it is absolutely FREE.

*DON'T SKIP THE LAB EXERCISE., IT IS HELPFUL*

This is the Learning Path

1) Introduction to Data Science
(https://lnkd.in/fF79bEj)

2) Data Science Tools
(https://lnkd.in/fYf2ZC8)

3) Data Science Methodology
(https://lnkd.in/fY6Kwqd)

4) Statistics 101
(https://lnkd.in/fpgJf7D)

5) Predictive Modeling Fundamentals I
(https://lnkd.in/f9_Y7UZ)

6) Python for Data Science
(https://lnkd.in/fy8E2wH)

7) Data Analysis with Python
(https://lnkd.in/fRQWByd)

8) Data Visualization with Python
(https://lnkd.in/fFu93ME)

9) Machine Learning with Python
(https://lnkd.in/f_7r534)

10) Deep Learning Fundamentals
(https://lnkd.in/fNvPvix)

11) Deep Learning with TensorFlow
(https://lnkd.in/ftfRtvQ)

#datavisualization #deeplearning #datascience #python #predictivemodeling #machinelearning

Please share us if you would like

✴️ @AI_Python_EN

Google Releases TensorFlow Federated, an Open-Source Framework to Facilitate Collaborative Machine Learning without Centralized Training Data

http://bit.ly/2EPsYy8

#MachineLearning #ArtificialIntelligence #DataScience

✴️ @AI_Python_EN

Do Neural Networks Need To Think Like Humans?

#neuralnetwork


https://www.youtube.com/watch?v=YFL-MI5xzgg

✴️ @AI_Python_EN

Marketing research almost seems to have an obsession with story telling which, perhaps, is telling. :-)

Seriously, I was at first taken aback when story telling began to get a lot of buzz because to me the word had negative connotations. (Read BS.)

Surely, there still are boring, disorganized presentations of research findings, e.g., hundreds of slides and perspiring presenters simply reading off numbers. How common this is is hard to tell.

Why does this happen? Inexperienced research execs with no time or budget to do their homework is one reason. Limited formal training in research is another. There surely are others.

Notice I've used the words why and cause. I think they are clues why some presentations fail and also keys to good storytelling (in the positive sense). We need to think causally. In fiction, events and characters are linked together causally. Even acts of nature have consequences.

Sometimes clients just want numbers or general reactions of focus group participants. At other times, though, they want us to help them better understand the why, not just the what, how, when, and so on.

Truly establishing causation is difficult to impossible even in "hard" science. But I think we often can do a better job without telling stories (in the negative sense).

Please share us if you would like

✴️ @AI_Python_EN

"OpenAI GPT-2: Understanding Language Generation through Visualization"

How the super-sized language model is able to finish your thoughts.

Blog by Jesse Vig: https://lnkd.in/ebHrTUP

#artificialintelligence #deeplearning #machinelearning

✴️ @AI_Python_EN

5 articles to learn #statistics for #datascience:

1. Comprehensive Inferential Statistics for Data Science -https://bit.ly/2NUQywr
2. Master Hypothesis Testing for Framing Data Science Problems - https://bit.ly/2u0utmV
3. Introduction to #ANOVA (with practical #Excel examples) - https://bit.ly/2F1ciE5
4. Tutorial for Understanding Non-Parametric Statistical Tests - https://bit.ly/2CcSxrr
5. Learn Statistics using R! - https://bit.ly/2VMOOIr

✴️ @AI_Python_EN

Do you remember a bullshit study published a few years ago claiming that deep learning can spot criminals from their photos and arguing that criminals have different facial features. Despite the ethical issue, we know this is bullshit but we couldn't spot the flaws.
Well, like most machine learning problems the devil is in the data.
To train the model the researchers used 700 of criminals ID photos as positive images. On other hands, they collected 1100 non-criminals from the web which featured people smiling.
No wonder why they go 90% accuracy!
So instead of developing criminals detector, they developed smiles detector LOL.
#research #machinelearning #deeplearning #ai
https://lnkd.in/fMhU4ZZ

✴️ @AI_Python_EN