DAY 1:
I have started with pandas which is a python library for data analysis.
DAY 2:
I brushed up my knowledge on numpy and practiced a few programs on HackerRank.
DAY 3:
I started browsing the basic difference in pandas python library.
DAY 4, 5: The tasks I have accomplished are understanding :
- cost functions
- gradient descent
- linear regression from scratch
DAY 6,7:
By the end of #week1ofmlcode in 100 days of ml coding challenge, I thought of taking up a course on Coursera -"Introduction to Data Science in Python" which would help me to emulate the best and I successfully completed the course which helped me in understanding data manipulation and cleaning techniques using the popular python pandas data science library.
Linkedin post
code is available.
DAY 8,9:
I have learned efficient ways to use the Mathplotlib which is a plotting library for the Python programming language and its numerical mathematics extension NumPy.
DAY 10,11:
LOGISTIC REGRESSION
Linkedin post
code is available.
DAY 12: understood the following concepts: fixing missingdata using 'Imputer', categorical data using 'LabelEncoder' and dummy Encoding using 'OneHotEncoder' and finally splitting dataset into 'testset' and 'trainset' using train_test_split from sklearn libraries.
DAY 13: gone through a few research papers which helped in understanding the real world applications of machine learning. youtube
DAY 14: DECISION TREE: Basic concepts and a few terminologies like prunning, information gain, gini impurity etc. link to reference.
DAY 15: Decision tree implementation from scratch using python. youtube
DAY 16:
Implementation of decision tree using sklearn.
Detailed explanation on linkedin
DAY 17: TOPIC: Probability UNDERSTANDING: In machine learning, the concept of probability plays an important role. For example, we ask our machine learning algorithm how likely our result is. Thus, after knowing the importance I started a playlist of Prof. Joe Biltzstein from Harvard University. Source
DAY 18:
Watched a few videos and read some articles on ml.
Source
DAY 19-23: Proper understanding on probability.Wrote my first blog article INTUITIONS ON PROBABILITY
DAY 24-26: Participated in my first real-world competition
DAY 27,28: Naive Bayes Classifier: This algorithm is mostly used in text classification/ Spam Filtering/ Sentiment Analysis/Recommendation System and with problems having multiple classes.Knew 3 different popular Naive Bayes classifiers.
DAY 29:
Naive Bayes
Infografic
DAY 30:
Naive bayes code from scratch using python
DAY 30-34: SVM article.
DAY 35-37:
Linear and non Linear classification using SVM:
Based on the dataset, if it is linearly separable LinearSVM can be used or
else kernel trick is used to project the data into a high-dimensional space before attempting to find a hyperplane.
Spam classification:
Many email services today provide spam filters that are able to classify emails
into spam and non-spam email with high accuracy.SVMs are used to build own spam filter.
DAY 38-40:
Ensemble learning helps improve machine learning results by combining several models. This approach allows the production of better predictive performance compared to a single model.
references:https://blog.statsbot.co/ensemble-learning-d1dcd548e936
Linkedin post
DAY 41:
Participated in an online hackathon
on data science
DAY 42:
By using Sklearn libraries implemented
Useful resources
DAY 43,44:
Flask is the best choice for all the web programmers, to start from the scratch and for building minimalistic web applications. One more advantage of using Flask is you can integrate machine learning algorithms(Python Based) within the functions with ease which makes your applications more intelligent and cognitive.
Here are few websites models you can build with Flask:
Blog Applications, Chat Applications, Data Visualisation, Dashboards, REST Applications, Admin Pages, Email-Services.
so I did a small Project on building a database for an art gallery
DAY 45,46:
As teaching gives a better learning, I've prepared and given a seminar at Gandhi Institute of Technology & Management (GITAM) University, Visakhapatnam today. I've introduced the standard ML problem types (classification and regression) and discuss prediction functions, feature extraction, learning algorithms, performance evaluation, cross-validation, sample bias, nonstationarity, overfitting, and hyperparameter tuning.
DAY 47-50:
Data Preprocessing:
https://www.youtube.com/watch?v=0xVqLJe9_CY&list=PL2-dafEMk2A64Er6yZHw1CJoFWQqYxj4B
https://www.analyticsvidhya.com/blog/2016/07/practical-guide-data-preprocessing-python-scikit-learn/
https://www.kdnuggets.com/2017/06/7-steps-mastering-data-preparation-python.html
DAY 51:
Introduction to Supervized learning
Clustering Algorithms (K-means): k-means clustering is a method of vector quantization, originally from signal processing, that is popular for cluster analysis in data mining. It has two main steps.
STEP ONE: Cluster assignment
STEP TWO: Move the centroid
DAY 52-57:
Pandas and Numpy
Prepared and gave a seminar
DAY 58:
As before starting to learn any new technologies, knowing its use cases and real world applications helps in improved learning,i have started with knowing the amazing applications of deel learning and how ks it different from machine learning
DAY 59-60:
Started course.fast.ai which is a top down approach of doing deep learning
DAY 61-100:
Finished one of the specializations of deeplearning.ai. And took pytorch course too
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent