60 DAYS OF UNMISSABLE UDACITY

60 DAILY DIGESTS OF THE REMARKABLE INTELLECTUAL ADVENTURE

The immensity of the Secure and Private AI subject is infinite. Every day of the #60daysofudacity had many questions of interest, but there was no day that did not reflect the community influence. I had been privileged to be a part of it. The Udacity community rendered help far beyond any expectations. For #60daysofudacity to happen, I had been dependent on other Udacity Scholars. It made all the difference between the steady learning progress and confusion. Many tireless Udacity Scholars provided me with invaluable assistance combined with priceless encouragements and countless ideas. My appreciation is boundless.

I'm Super Excited, Let's Get Started

Day 1, June 28th, Friday:
- Prioritising & Planning
  
  Secure and Private AI Challenge content summary:
  - Differential Privacy
  - Evaluating the Privacy of a Function
  - Local and Global Differential Privacy
  - Differential Privacy for Deep Learning
  - Federated Learning
  - Securing Federated Learning
  - Encrypted Deep Learning
- Started from Differential Privacy and coded the examples.
- Let's see how I can do that. Link to GitHub repo:
```
   https://github.com/reallygooday/60daysofudacity/blob/master/Differential_Privacy.ipynb
```

Day 2, June 29th, Saturday: Gaining intuitive understanding about the basics of Neural Networks:
- First Takeaway --> the term 'Logistic Regression' represents a small Neural Network and learning algorithm for Binary Classification Problem;
- Second Takeaway --> the loss function of Logistic Regression computes the error for a single training example, the cost function of Logistic Regression is the average of the loss function of the entire training set);
- Third Takeaway --> the term 'Gradient Descent algorithm' represents the slope of the function;
- Forth Takeaway --> computations of Neural Networks organised in terms of a forward pass (or a forward propagation step), in which we compute output of Neural Networks, followed by a backward pass (or a back propagation step), which we use to compute gradients or derivatives.

Day 3, June 30th, Sunday:
- doing the programming exercises;
- learning about vectorization, re-writting explicit 'for' loops;
- non-vectorize, looped code takes much longer than vectorize code;
- read the article 'Why you should forget ‘for-loop’ for data science code and embrace vectorization'
- link to the article:
```
           https://towardsdatascience.com/why-you-should-forget-for-loop-for-data-science-code-and-embrace-vectorization-696632622d5f   
```
- The Thing To Remember: Vectorize Your Code to speed up, to get results faster !

Day 4, July 1th, Monday:
- read the article 'Broadcasting with PyTorch';
- I allude briefly to the main points, Rules of Broadcasting -->
  - Rule 0: Each tensor has at least one dimension(trivial)
  - Rule 1: If the two arrays differ in their number of dimensions, the shape of the one with fewer dimensions is padded with ones on its leading (left) side.
  - Rule 2: If the shape of the two arrays does not match in any dimension, the array with shape equal to 1 in that dimension is stretched to match the other shape.
  - Rule 3: If in any dimension the sizes disagree and neither is equal to 1, an error is raised.
- link to the article:
```
           https://mc.ai/broadcasting-with-pytorch/   
```

Day 5, July 2th, Tuesday: learned:

how to authenticate Google Colab access to Google Drive and mount Google Drive

                         from google.colab import drive
                         drive.mount('/content/gdrive')

how to save the trained model to Google Drive with torch.save;
how to load the model from Google Drive and reuse it with torch.load.

Link to the article with more details:

                         https://medium.com/@ml_kid/how-to-save-our-model-to-google-drive-and-reuse-it-2c1028058cb2

Day 6, July 3th, Wednesday:

worked on Credit Risk project using Logistic Regression Model,

link to the project:

            https://github.com/reallygooday/60daysofudacity/blob/master/CreditRisk.ipynb

started Udacity "Introduction to Deep Learning with PyTorch course",

watched the following video-lessons:

           - Introduction to Neural Networks,
          
           - Convolutional Neural Networks,
           
           - Style Transfer,
           
           - Recurrent Neural Networks,
           
           - Sentiment Prediction RNN,
           
           - Deploying PyTorch Models.

Day 7, July 4th, Thursday:
- watched Lesson 2 'Deep Learning with PyTorch',
- learned about Network Architectures, how to accelerate network computations using a GPU, how to implement Softmax.

Day 8, July 5th, Friday:
- watched Lesson 3 'Introducing Differential Privacy',
- learned about basics of differential privacy, a method for measuring how operations impact the privacy of data.

Day 9, July 6th, Saturday:
- watched Lesson 4 'Evaluating Privacy of a Function',
- implemented differential privacy.

Day 10, July 7th, Sunday:
- watched Lesson 5 'Introducing Local and Global Differential Privacy',
- learned how to apply differential privacy to arbitrary algorithms by adding noise to the outputs,
- getting comfortable coding with tensors,
- link to GitHUb Jupyter Notebook:
```
         https://github.com/reallygooday/60daysofudacity/blob/master/intro_to_tensors.ipynb
```
- link to Colab Jupyter Notebook:
```
         https://colab.research.google.com/drive/1Z2o5t8-yrwtu9LCRgcbxmCwTyaQOHx-s#scrollTo=hI7zQLnAf1Q3  
```

Day 11, July 8th, Monday:
- It took 1 hour to train a basic deep learning chatbot model on my laptop. Then I moved the model to Colaboratory (Colab, Google free Cloud service), adjusted settings, selected GPU as Hardware accelerator, mounted my Google Drive, downloaded the dataset (corpus) directly to Google Drive via Google Colab. It took a few minutes to run the whole model with Colab. Nothing gets me moving like a completed task.
```
- Link to GitHub Jupyter Notebook: 

           https://github.com/reallygooday/60daysofudacity/blob/master/chatbot.ipynb

- Link to Colab Jupyter Notebook: 

           https://colab.research.google.com/drive/1XgymVaAdjrEuMWGxekqlMvtgkSBLZLIz#scrollTo=72Udiv_SZIzO
```

Day 12, July 9th, Tuesday:
- watched Lesson 6 'Differential Privacy for Deep Learning';
- learned how to apply differential privacy to deep neural networks;
- participated in study group #sg_applied_dl virtual discussion from 3:30pm on Slack;
- worked on the chatbot application;
- started reading book "PyTorch A Problem-Solution Approach" by Pradeepta Mishra.
- Link to Slack post:
```
          https://github.com/reallygooday/60daysofudacity/blob/master/day_12.pdf
```

Day 13, July 10th, Wednesday:
- watched Lesson 7 'Federated Learning',
- learned about FL, a method for preserving data privacy by training models where the data lives,
- worked on Kaggle Dog Identification project,
- encouraged @Ricardo Felipe Redin @Eric Ndirangu @Ahmad Waly @Y.Leonce Eyog @PJha to share the progress on learning and @Marcos Junior @Fitsum Ayalew to take up the challenge.

Day 14, July 11, Thursday:
- watched Lesson 8 'Securing Federated Learning',
- learned about secure models trained using Federated Learning with multi-party computation,
- continued working on Kaggle Dog Breed Identification project,
- encouraged @hammedb @Aakanksha Dhurandhar @Idan Benaun to share the learning progress.

Day 15, July 12, Friday:
- watched Lesson 9 'Encrypted Deep Learning',
- learned how to perform encrypted computation, builded on encrypted database, and generated an encrypted prediction with an encrypted neural network on an encrypted dataset,
- continued working on Kaggle Dog Breed Identification.
- encouraged @Pablo Carmona - Zuluaga @hammedb @Idan Benaun @Idan Benaun @uday.kiran @Yanbin Zhang @Temitope Oladokun @Rabin Banjade @Elif Yalı @R Aishwarya @Elif Yalı @Waad Gamal Shehabeldin @qayyumb to share the learning progress. Time, when it left to itself, cannot be trusted to move at any pace.

Day 16, July 13, Saturday:
- Continued working on Dog Breed Identification project.
- Encouraged @Lakshay Wadhwa @K3NNi @J.D.V.P.REDDY @Hachiman @Gabor Soter @Fabio @Daibin Raju @Waad Gamal Shehabeldin @R Aishwarya @medhagupta098 @falconArrow @hammedb @Vaibhav @Claudia Hill @Xiaolei Zhang @Yanbin Zhang.

Day 17, July 14, Sunday:
- Continued working on Dog Breed Identification project

Day 18, July 15, Monday:
- Registered for AWS Online Conference on the 18th July to learn how to deploy a basic deep learning model on cloud;
- Encoraged @ Marcos Junior to take up challenge;
- Continued working on Dog Breed Identification project on Kaggle (got stuck with Value Error)
- Link to the Kaggle kernel:
```
https://www.kaggle.com/reallygooday/dog-breed-identification-exploration
```

Day 19, July 16, Tuesday:
- Read book "PyTorch Recipes" by Pradeepta Mishra.

Day 20, July 17, Wednesday:
- Read book "Deep Learning with PyTorch" by Vishnu Subramanian;
- Registered for AWS DeepRacer Challenge with Udacity.

Day 21, July 18, Thursday:
- Attended/ watched the AWS Innovate Conference online to learn about PyTorch inference with Amazon SageMaker.
- Link to the presentation file:
```
 https://github.com/reallygooday/60daysofudacity/blob/master/PyTorch_Opensource%2BML%2Bframeworks%2Bon%2BAmazon%2BSageMaker.pdf
```

Day 22, July 19, Friday:
- Read the book "Deep Learning with PyTorch" by Vishu Subramanian.
- Link to Slack post:
```
https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1563570031267300
```

Day 23, July 20, Saturday:
- Keep on reading the book "Python Data Analytics" Fabio Nelli.
- Encouraged @Qazi Rameez Bin Naeem @Ahmad Waly @Eben O @Reem Qais @Tasmiyah Qazi @Elif Yalı@Ugesh S @hamzzy @Temitope Oladokun @Pablo @R Aishwarya @Saleeta Qadir@Daibin Raju @Fabian Schimpf @Gaurav Yadav @Heather Allen @Jacqueline @hamzzy@Jeremiah Kamama @karthik s @Lakshya Mutneja @Cristina Zapata @Jia Xia @Christian I. Muñoz @Vaishnav Potlapalli @Bharat Bhushan Mishra @Nada Hammad @Superuser@Udolf to share your learning adventure and progress.
- Link to Slack post:
```
https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1563659304247700
```

Day 24, July 21, Sunday:
- Reading the book "Python Deep Learning" (exploring deep learning techniques and neural network architectures with PyTorch) by Ivan Vasilev, Daniel Slater, Gianmario Spacagna, Peter Roelants, Valentino Zocca.
- Encouraged @Reem Qais @Cristian Lazo Quispe @Brandon W @Eman @Nada Hammad @Ravikant Singh @Tesin Dosch @Ayush Yadav @Udit Singhania @Ignacio Palos Reynoso @Christian I. Muñoz @Olivia Wais @Pablo Martínez @R Aishwarya @Fabian Schimpf @Dheemanth M D @Faiz Kidwai @Felipe Damascena @Gabriel Giro Resende @Hafsa Sam @Jade Mays @Hadeel Alabandi @K3NNi @kumarklc to keep constant thirst for new knowledge.
- Link to Slack post:
```
https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1563745328450600
```

Day 25, July 22, Monday:
- Getting comfortable coding and making predictions with a basic PyTorch Linear Model.
- Link to Colab Jupyter Notebook:
```
https://colab.research.google.com/drive/19QYWp_lExrN65b6OesWS9NNA7x0fHKmq#scrollTo=KOl9mFXs7ja8----
```
- Link to GitHUb repo Jupyter Notebook:
```
https://github.com/reallygooday/60daysofudacity/blob/master/Day25_Linear_Model.ipynb
```
- Encouraged to succeed the following talented students @hamzzy @Nabilah Hannani @R Aishwarya @birozso @Pablo @Naveen Tadala @Vaibhav @Claudia Hill @Xenofon Karamanos @zilu.
- Link to Slack post:
```
https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1563809544107500    
```

Day 26, July 23, Tuesday:

Trained a linear model to fit training data using gradient descent.

Link to GitHub Jupyter Notebook:

 https://github.com/reallygooday/60daysofudacity/blob/master/Day26_July22_Linear_Regression.ipynb

Link to Colab Jupyter Notebook:

 https://colab.research.google.com/drive/19QYWp_lExrN65b6OesWS9NNA7x0fHKmq#scrollTo=0-bLHYJDxUzn

Encouragements to @Saleeta Qadir and @waleed hassan.

Link to Slack post:

 https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1563875855292000?thread_ts=1563875855.292000

Day 27, July 24, Wednesday:

✔️ Started reading book "Python Data Science Essentials " by Alberto Boschetti and Luca Massaron;

✔️GitHub repo:
```
    https://github.com/reallygooday/60daysofudacity;
```
✔️ Encouraged awesome people @Sankalp Dayal @Evi @Archit @Berenice @Maria Medina @Nabilah Hannani @Brendan Odigwe @AaronStone @Christine Rehm @Biswajit Banerjee @qayyumb to have audacious goals. :bettertogether:

Link to Slack post:
```
    https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564009106372700   
```

Day 28, July 25, Thursday:

✔️Getting comfortable building a basic image classification;

✔️GitHub Jupyter Notebook:

      https://github.com/reallygooday/60daysofudacity/blob/master/Basic_Image_Classifier.ipynb

✔️ Link to Slack post:

      https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564095577204600

Day 29, July 26, Friday:

Read article “How to Use Deep Learning to Write Shakespeare”

   https://opendatascience.com/how-to-use-deep-learning-to-write-shakespeare/;

Read book "Python Data Science Handbook" by Jake Vander Plas

   https://jakevdp.github.io/PythonDataScienceHandbook/.

Link to Slack post:

   https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564182392005300

Day 30, July 27, Saturday:

Experimented with Style Transfer, stylised photos with neural networks using PyTorch.
Encouraged @Nabilah Hannani @Elif Yalı @qayyumb @AaronStone @Reem Qais @Ayush Yadav to stay bold and Udacious !

Link to Slack post:

    https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564256061078300

Link to Github repo:

                 https://github.com/reallygooday/60daysofudacity/blob/master/Claude_Monet.ipynb   
                 https://github.com/reallygooday/60daysofudacity/blob/master/Leonardo_di_ser_Piero_da_Vinci.ipynb

Day 31, July 28, Sunday:
- Experimented with Image Classifier.
- Link to Github Jupyter Notebook: Image Classifier :
  - https://github.com/reallygooday/60daysofudacity/blob/master/Image_Classifier.ipynb
- Link to Slack post :
  - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564353644124500

Day 32, July 29, Monday:

Experimented building Text Generation with PyTorch using the book "Oliver Twist" by Charles Dickens.

Link to the book:

          http://www.gutenberg.org/cache/epub/730/pg730.txt.

Link to the tutorial:

          https://machinetalk.org/2019/02/08/text-generation-with-pytorch/

Link to Slack post:

          https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564414980191500

Day 33, July 30, Tuesday:
- Experimented building Story Generator with PyTorch using the story "A Clean, Well-Lighted Place" by Earnest Hemingway.
- Revised Introduction to Deep Learning Lesson 7 "Character-Level RNN, text prediction with an LSTM".
- Link to Slack post:
```
           - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564513286344500  
```
- Link to Github repo:
```
           - https://github.com/reallygooday/60daysofudacity/blob/master/Hemingway.ipynb   
```

Day 34, July 31, Wednesday:
- Restyled images with Neural Network using personal photos as content.
- Submitted the restyled images for Project T-Shirt.
- Encouraged @Elif Yalı @Saleeta Qadir @waleed hassan @Elif Yalı @Rabin Banjade @Tasmiyah Qazi @Y. Leonce Eyog @Udolf @Idan Benaun @Olivia Wais @Pablo Martínez @Udolf @Saleeta Qadir @Daibin Raju @Faiz Kidwai @Faiz Kidwai @Gabor Soter @hamzzy @Jade Mays @Jeremiah Kamama @Luis Bronchal. Happy Pytorching ! :pytorch:
- Link to Slack posts:
```
           - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564589848270300                    
           - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564589676266500
```
- Link to Facebook Project T-Shirt submission page:
```
           - https://www.facebook.com/media/set/?set=oa.812067435861177&type=3         
```

Day 35, August 1, Thursday:
- Started Cloud Qwiklabs and deeplearning.ai courses; I'm interested in Kubernetes in the Google Cloud.
- Link to Slack posts:
```
             - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564698476117700   
```

Day 36, August 2, Friday:

Revised Differential Privacy Lessons, building a private database.

Link to Slack post:

             - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564779393239500

Day 37, August 3, Saturday:
- 🚙 Started with AWS DeepRacer and the AWS platform training autonomous racer using reinforcement learning.
- 📚 Read the book “Differential Privacy and Applications” by Tianging Zhu, Gang Li, Wanlei Zhou, Philip S.Yu
- 🙂 Encouraged @Maria Medina @Nada Hammad @birozso @Vaishnav Potlapalli @Çağatay Onur Şengör @Xiaolei Zhang @Zain Sarwar @Lakshya Mutneja @Xenofon Karamanos @Jeremiah Kamama at all times, particularly in times of uncertainty, to try your absolute hardest to keep going forward learning about Private AI and to go 'all-in'. Wishing you loots of Deep Learning fun ! Happy times to you all !
- Link to Slack post:
```
              - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564864391377600 
              
              - https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564851898467400 
```

Day 38, August 4, Sunday:
- worked on programming exercises related to Broadcasting with PyTorch and Vectorisation.
- Link to Slack post:
```
       https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1564952806091900  
```

Day 39, August 5, Monday:

read the paper about MSG-Net; link to the paper:

      https://github.com/reallygooday/60daysofudacity/blob/master/academic%20paper/Zhang_Multi-style_Generative_Network_for_Real-time_Transfer_ECCVW_2018_paper.pdf

experimented with PyTorch MSG-Net:

   The Multi-style Generative Network is the first to achieve real-time brush size control for multi-style  transfer.
    The picture below shows the Style Transfer results. Personal photos used as content.

Encouraged @Mansi Goyal @MasterKN @BMC @Vaibhav @Pablo Carmona - Zuluaga @Xiao O @Cristina Zapata @Lisa M. Venezia @Kabiswor Joshi @Jacob Pardue @Hafsa Sam @Gabriel Jiménez Forero @Fabliha Fairooz @Dinak Lal B . You are awesome ! 😃

Link to Slack post:

   https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565044485292700    
   https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565034137147700

Day 40, August 6, Tuesday:
- Investigated, worked on correcting the following errors: import error in coded implementation of encrypted training with PySyft’, index error in coded implementation of text generator with character level RNN.
- Encouraged the following wonderful scholars @Jeremiah Kamama, @Tasmiyah Qazi @Chetto13 @Wamaitha @Eko Edita Limanta @Rachesh @Hung @Noom Chanthadit @Ujjwal @ibasly @Octavio @Pooja Vinod to push the limits of own superpower.
- Link to Slack post:
```
       https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565118650143600
```

Day 41, August 7, Wednesday:
- getting back to basics, continued coding activities using tutorials for beginners in PySyft (https://github.com/OpenMined/PySyft/tree/dev/examples/tutorials).
- Encouraged the following amazing scholars @xworow @Caitlyn Singam @Vaibhav Sachdeva @bhavz @Noom Chanthadit @Mahi @Larissa de Moura @Jia Xia @medhagupta098 to act one step ahead in learning.
- Link to Slack post:
```
       https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565214285386900               
```

Day 42, August 8, Thursday:
- continued learning about RNN (both GRU and LSTM) from Coursera - specific types of RNN architecture with sequence data (many-to-one, many-to many, one-to-many);
- participated in #jobs channel activity Resume Evaluation.
- Encouraged @Vaibhav @Nachiket @Fahad Farooq @Hafsa Sam @Aim
- Link to Slack post:
```
       https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565304435463700               
```

Day 43, August 9, Friday:
- read the following research paper : "Representation Learning with Contrastive Predictive Coding" about unsupervised learning approach of extracting representations from data, the proposed approach can be used in speech, images, text and reinforcement learning.
- Link to the research paper:
```
  https://arxiv.org/pdf/1807.03748.pdf  
```
- Encouraged the following amazing scholars @B V Pradeep @Lakshmi @kumarklc @AaronStone @sunnets @Dheemanth M D @Fabio Valerio Massoli @Gabor Soter to think about your Future Self and to make Future-Friendly choices, to smarten yourself up with the right balance of learning effort and progress.
- Link to Slack post:
```
 https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565391347473400
```

Day 44, August 10, Saturday:
- connected a Postgres database running on Panoply to Jupyter Notebook;
- analysed PostgresData in Jupyter Notebook.
- Link to Slack post:
```
  https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565477346469100
```

Day 45, August 11, Sunday:
- experimented coding threading Lock for Parallel Processing.
- Encourage the following well known Udacians @Abhimanyu @Berenice @Sankalp Dayal @Evi @Frida @Anita Goldpergel @Dharmendra Choudhary @Cristina Zapata @Sejas @Ahmed Thabit. Nothing worthwhile is ever achieved without the consistent effort. True greatness only comes to those willing to do the routine, daily work. Forgive me for the stating the obvious, you're doing a fantastic job sharing the memorable experiences every day !
- Link to Slack post:
```
           https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565563557002600  
```

Day 46, August 12, Monday:
- learned about recursion, recursive algorithms and the time complexity of linked list operations.
- Encouraged the following wonderful scholars @zack @xworow @Çağatay Onur Şengör @Vebby @Baran Berkay Barakçin @Nana Aba T @Mahitej to share daily updates, knowledge and ideas.
- Link to Slack post:
```
           https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565649865430800 
```

Day 47, August 13, Tuesday:
- worked on federated learning with encrypted gradient aggregation.
- Encouraged the following wonderful scholars @R Aishwarya @Hafsa Sam @Daibin Raju @Fabrice @medhagupta098 @Nabilah Hannani to share your learning progress.
- Link to Slack post:
```
           https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565736159316600
```

Day 48, August 14, Wednesday:

revised PATE Analysis and Differential Privacy for Deep Learning.

Link to Slack post:

           https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565822002232800

Day 49, August 15, Thursday:
- revised Differential Privacy (Implementing Local Differential Privacy, Varying the Amount of Noise, Creating a Differentially Private Query);
- read research paper "Selective Classification for Deep Neural Network" by Yonatan Geifman, Ran El-Yaniv; in this paper, a method to construct a selective classifier given a trained neural network, is proposed; Selective Prediction is a subfield of Machine Learning dealing with the ability to know what you don't know, which can make you smarter; link to the paper:
```
https://papers.nips.cc/paper/7073-selective-classification-for-deep-neural-networks.pdf.
```
- Encouraged the following wonderful scholars @Zain Sarwar @Jia Xia @Caitlyn Singam @Vebby @bhavz @Thor @Jacqueline @Jeremiah Kamama @Lala @Mahak Bansal @B V Pradeep @Oge @ibasly @Ujjwal @Yanbin Zhang @Noom Chanthadit @Rabin Banjade @rajeman @Wamaitha @Qazi Rameez Bin Naeem .
- Link to Slack post:
```
  https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565909540141900  
```

Day 50, August 16, Friday:

worked on the Text Generator project.

Link to Slack post:

   https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1565987525407100

Day 51, August 17, Saturday:
experimented using Recursion and Advanced Data Structures (Linked Lists and Hash Tables);
experimented using cryptographic hash function to detect data corruption errors and verify data integrity
```
     (http://blog.chapagain.com.np/cryptography-verify-file-integrity-with-md5-hash-checksum/).
```
encouraged the following great scholars @birozso @xworow @Noom Chanthadit @Vaishnav Potlapalli @Rabin Banjade @Naveen Tadala @Mahitej @Nana Aba T and invited to share takeaways from what they learned.

Link to Slack post:

                    https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566043025313000

Day 52, August 18, Sunday:
- read research paper "Deep Convolutional Neural Networks for Sentiment Analysis of Short Text"
- link to the paper:
```
                 https://www.aclweb.org/anthology/C14-1008 ;
```
- the main takeaways of the paper are - the idea of using CNN to extract from character-to sentence-level features and the demonstration that a feed-forward NN architecture can be as effective as RNTN for sentiment analysis of sentences.
- Encouraged the following wonderful scholars @Reem Qais @Waad Gamal Shehabeldin @Eko Edita Limanta @Qazi Rameez Bin Naeem @Rachesh @Nana Aba T @Y. Leonce Eyog @uday.kiran @Idan Benaun @Eben O @Pablo Carmona - Zuluaga @R Aishwarya @Fabian Schimpf @Daibin Raju @Fabliha Fairooz @medhagupta098 @Hung @AaronStone @bhavz @bhavz @Laura W .
- Link to Slack post:
```
               https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566148484356900 
```

Day 53, August 19, Monday:

read paper "Simplifying Graph Convolutional Networks";

link to the paper:

                 https://arxiv.org/pdf/1902.07153.pdf.

Link to Slack post:

                 https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566236184191600

Day 54, August 20, Tuesday:
- read paper Frankle, J., & Carbin, M. (2018). The lottery ticket hypothesis: Finding sparse, trainable neural networks. arXiv preprint arXiv:1803.03635;
- link to the paper:
```
        https://arxiv.org/pdf/1803.03635.pdf  
```
- Link to Slack post:
```
        https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566336315269300  
```

Day 55, August 21, Wednesday:
- read paper --> [18/07/2019] Gary presents: Zhang, G., Wang, C., Xu, B., & Grosse, R. (2018). Three mechanisms of weight decay regularization. arXiv preprint arXiv:1810.12281;
- link to the paper:
```
        https://openreview.net/pdf?id=B1lz-3Rct7  
```
- Link to Slack post:
```
        https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566424894430300  
```

Day 56, August 22, Thursday:
- read about No Free Lunch Theorem in Machine Learning (the theorem implies that a model that explains a certain situation well may fail in another situation);
- read paper [18/04/2019] Vahan presents: Vyas, A., Jammalamadaka, N., Zhu, X., Das, D., Kaul, B., & Willke, T. L. (2018). Out-of-distribution detection using an ensemble of self supervised leave-out classifiers. In Proceedings of the European Conference on Computer Vision (ECCV) (pp. 550-564)
- link to the paper
```
      http://openaccess.thecvf.com/content_ECCV_2018/papers/Apoorv_Vyas_Out-of-Distribution_Detection_Using_ECCV_2018_paper.pdf
```
- in this paper, the authors proposed an out-of-distribution detection method for deep neural networks and demonstrated its performance across several out-of-distribution classification tasks on the deep neural networks such as DenseNet[8] and Wide ResNet(WRN).
- Encouraged the following amazing scholars @Reem Qais @Waad Gamal Shehabeldin @Eko Edita Limanta @Qazi Rameez Bin Naeem @Rachesh @Nana Aba T @Y. Leonce Eyog @uday.kiran @Idan Benaun @Eben O @Pablo Carmona - Zuluaga @R Aishwarya @Fabian Schimpf @Daibin Raju @Fabliha Fairooz @medhagupta098 @Hung @AaronStone @bhavz @bhavz @Laura W to share your daily digest of learning activities with others.
- Link to Slack post:
```
     https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566512206458500  
```

Day 57, August 23, Friday:
- daily digest: learned about empirical and theoretical probabilities, the fundamental probability rules and how to solve complex probability problems using special counting techniques like permutations and combinations;
- takeaway: permutations are for lists (order matters) and combinations are for groups (order doesn't matter);
- example: there are 10,000 possible permutations for a 4-digit PIN code (there are 10,000 digit arrangements where the order of the digits matters);
- joke: a "combination lock" should really be called a "permutation lock".
- Encouraged the following wonderful scholars @hamzzy @Quan @Olivia Wais @Chetto13 @Rachel Beldner @T @Saleeta Qadir @Youmna Ebrahim @Ujjwal @Ibrahim Salihu Yusuf @Peagol @Pava to get the most out of your learning.
- Link to Slack post:
```
           https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566524610070500  
```

Day 58, August 24, Saturday:
- experimented with image classifier using VGG16 Model;
- link to Colab Jupyter Notebook:
```
   https://github.com/reallygooday/60daysofudacity/blob/master/Gender_Prediction.ipynb  
```
- read article 'Optimizing Jupyter Notebook: Tips, Tricks, and nbextensions'
- link to the article:
```
 https://towardsdatascience.com/optimizing-jupyter-notebook-tips-tricks-and-nbextensions-26d75d502663);
```
- read paper 'Pragmatically Informative Text Generation'
- link to the paper:
```
 https://arxiv.org/pdf/1904.01301.pdf).  
```
- Encouraged the following wonderful scholars @Quan @Olivia Wais @Edson Cavalcanti @Ramón Romero @T @Amresh Yadav @Ugesh S @Ibrahim Salihu Yusuf @Brendan Odigwe @Pava @Aim @sunnets @Doaa Mohammad Isa Odat @Fanni Ikhwan @Gabir Yusuf @Hafsa Ansari @Jinyi Tan not to get seduced by the idyllic weekend in the sun and to continue your productive learning.
- Link to Slack post:
```
       https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566605067338500  
```

Day 59 August 25, Sunday:
- daily digest: read Cornell Tech research papers about misuse, leaks of Private Data and how data privacy is evolving;
- link to the list of research papers: http://www.cs.cornell.edu/~shmat/research.html ;
- takeaway: Vitaly Shmatikov, professor of CS at Cornell Tech, developed tools that can indicate if your data was misused;
- jokes: ...before I tell you my name, I'll need to know how you're planning to use it....
```
       ...camera, smartphone, smartwatch, smart glasses, smart TV, tracking device, I now have everything except privacy...;  
```
- Encouraged the following wonderful scholars @Jacqueline @Pablo Carmona - Zuluaga @Quan @Brandon W @Abbie Weisenbloom @Ebtehal Khaled Seddiek @Rachesh @Tahir Khalil @Yassen Mehrez @Ugur Caliskan @Ibukun @uday.kiran @Idan Benaun @PJha @R Aishwarya @Saleeta Qadir @Daibin Raju @Faiz Kidwai @medhagupta098 @Gabor Soter @Hafsa Sam @Jade Mays;
- Link to Slack post:
```
          https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566738011007800  
```

Day 60, August 26, Monday:
- practiced coding Transfer Learning with VGGNet;
- practiced coding Parallel Processing (running multiple instructions at the same time );
- read more about Transfer Learning from the Stanford course notes here http://cs231n.github.io/transfer-learning/;
- link to the open Stanford course: http://cs231n.github.io/ ;
- read Andrei Karpathy blog "Hacker's guide to Neural Networks" http://karpathy.github.io/neuralnets/;
- Encouraged the following wonderful scholars
  @qayyumb @Lakshay Wadhwa @Chetto13 @Ebude @Rachel Beldner @Tadewos G. Tadewos @Yassen Mehrez @Ugur Caliskan @Ibukun @Oge @Padhma M @Aim @Ugesh S @Aim @Pakeeza. Keep up the amazing work !
- Link to Slack post:
```
                https://secureprivataischolar.slack.com/archives/CKRQP9ZLL/p1566858755322000  
```

divyawadehra / 60daysofudacity Goto Github PK

60daysofudacity's Introduction

60 DAYS OF UNMISSABLE UDACITY

60 DAILY DIGESTS OF THE REMARKABLE INTELLECTUAL ADVENTURE

I'm Super Excited, Let's Get Started

60daysofudacity's People

Contributors

Watchers

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent