Today's Work : I started course on 'Data Visualization With Python' on coursera. Completed Week-1 of it, wrangled dataset and visualized it on Line Chart.
Today's Work : Week 2 in 'Data Visualization With Python' on coursera, visualized data using area plots and other.
Today's Work : Completed Week 2 in 'Data Visualization With Python' on coursera, visualized data using box plot, scatter plot and bubble plot
Today's Work : Week 3 in 'Data Visualization With Python' on coursera. Advanced visualization of data using Waffle charts and Word Clouds. Explored Seaborn library and made regression plots with it. Visualizing Geospatial Data using Folium and Choropleth Maps. Also, completed 2/4 tasks of week 3 final assignment.
Today's Work :Earned Certification In 'Data Visualization With Python' By IBM on coursera. Completed peer-graded assignment comprising of 4 tasks. Started Course 8 (Machine Learning With Python) of 9 in the IBM Data Science Professional Certificate Specialization.
Today's Work : Completed Week 1 and started Week 2 In 'Machine Learning With Python' By IBM on Coursera, learned about Simple Linear Regression, Model Evaluation In Linear Regression and Evaluation Metrics in Linear Regression. Implemented Simple Linear Regression on predicting the CO2 Emission of Cars using Sci-kit learn and Visualized the Regression Fit Line.
Today's Work : Completed Week 2 In 'Machine Learning With Python' By IBM on Coursera, learned about Multiple Linear Regression, Polynomial Linear Regression and Non-Linear Regression. Implemented Multiple, Polynomial and Non-Linear Regression, used the Sigmoid/Logistic model to fit the data and found the Optimized Parameter using Scipy's Optimization Library (Curve_Fit for non-linear).
Today's Work : Week 3 In 'Machine Learning With Python' By IBM on Coursera, learned about Classification, types of Classification, got to know about different Models used for Classfication. Learned about K-Nearest Neighbor Model for Classification, implemented Multiclass Classification Problem for a Telecommunication company using KNN. Ran the model for different values of 'K' and picked the Optimal one on the basisc of Accuracy Score (Sort Of Jaccard Index Score) for each 'K', computed the Standard Deviation for each Accuracy Score. Plotted the result Accuracy Score values with respect to Standard Deviation for each Score.
Today's Work : Week 3 In 'Machine Learning With Python' By IBM on Coursera. Learned about Decision Tree Model in Classification and Implemented it on Binary Classification problem for Drug Recommendation.
Today's Work : Week 3 In 'Machine Learning With Python' By IBM on Coursera. Learned about Logistic Regression for Classification. Learned about Metrics used for Evaluation of the model like Confusion-Matrix, Log-Loss, Jaccard-Similarity-Score etc. Learned about the parameter 'C' and 'Solver' worked with differnet values for 'C' and 'Solver'. Implemented Logistic Regression on a hypothetical Communication Company named Churn by using the scikit leaern LogisticRegression model for C=1 and solver='newton-cg' whereas the model was efficient for parameters C=0.01 and solver='liblinear' which produced logloss 0.60 and 0.25 respectively.
Today's Work : Week 3 Completed In 'Machine Learning With Python' By IBM on Coursera. Learned about Support Vector Machine (SVM) model for Classification. Learned about what 'Kernel' is and impletemented SVM model for 2 different Kernels. This Classification model was impletementd on data about Patients being Benign or Malignant.
Today's Work : Week 4 In 'Machine Learning With Python' By IBM on Coursera. Intorduction to Clustering and different types of Clustering. Understanding about Clustering as Unsupervised Machine Learning technique. Leaned about K-Means Clustering Algorithm that falls in Partion Clustering type. Implemented K-Means Clustering model on random data and then for Customer Segmentation based on their Demographical Data.
Today's Work : Week 4 Completed In 'Machine Learning With Python' By IBM on Coursera. Learned about Hierarchical Clustering and its pros and cons in over KMeans( Partition Clustering). Used scipy library for Dendogram visualization. Found out the Optimal number of Clusters by looking at the Dendogram. Implemented Heirarchical Clustering using scipy and scikit libraries. Learned about another Clustering technique Density Based Spatial Clustering of Applications With Noise (DBSCAN). Implemented DBSCAN for Clustering Weather Stations based on their Location, Mean, Min and Max temperature.
Today's Work : Week 5 Completed In 'Machine Learning With Python' By IBM on Coursera. Learned about Recommendation System. Worked on Content-based Recommender System for recommending Movies to the User based on his ratings given to the movies he has watched. Learned about Collaborative filtering. Used Pearson Coefficient for similarity measure of the Users from the target User, hands on lab for Collaborative Filtering.
Today's Work : Week 6 In 'Machine Learning With Python' By IBM on Coursera. Started Working on Final Project of the course i.e. Making a Best Classifier Model for Predicting if a Person can take Loan from a Bank or not.
Today's Work : Week 6 Cont'd In 'Machine Learning With Python' By IBM on Coursera. Working on Final Project of the course i.e. Making a Best Classifier Model for Predicting if a Person can take Loan from the Bank or not.
Today's Work : Week 6 Completed In 'Machine Learning With Python' By IBM on Coursera. Final Project of the course (The Best Classifier) completed. Started exploring Deep Learning. Worked with Artificial Neural Network, learned about Feedforward and Back Propagation, learned about Batch Gradient Descent and Stochastic Gradient Descent, used Keras Library with Tensorflow backend to make an Aritificial Neural Network for Bank to find the Probability about Person leaving the Bank. Compared the ANN's accuracy with simple scikit learn's LogisticRegression model accuracy. ANN was more accurate than simple Machine Learning LogisticRegression model.
Today's Work : Started working on Convolutional Neural Network, learned about each step involved in Convolutional Neural Network. Leaned about extracting features through Convolution, then by using Max Pooling technique to get Critical Details of the feature. Then, applied Flattening algorithm and made a Full-Connection. This CNN is used for predicting if the the Image is of a Cat or a Dog.
Today's Work : Convolutional Neural Network Completed, made a CNN for Image Classification for 2 Subjects (Cat and Dogs) with Accuracy of 89%.
Today's Work : Introduction to Dimensionality Reduction. Worked with Principal Component Analysis for Dimensionality Reduction.
Today's Work : Dimensionality Reduction - Worked with Linear Discriminant Analysis for Dimensionality Reduction.
Today's Work : Dimensionality Reduction - Worked with Kernel Principial Component Analysis (Kernel PCA) for Dimensionality Reduction.
Today's Work : Model Selection - Worked with K-Fold Cross Validation for Model Selection, also learned about finding the best Hyperparameters for Model using Grid Search. Learned about XGBoost.
Today's Work : Introduction to Natural Language Processing (NLP) - Worked with NLP for a Restaurant. Made a Classification Model based on Reviews given to Restaurant by Customers to Predict if the Customer liked the Restaurant or Not.
Today's Work : Started Working on Document Clustering. Learned about Term-Frequency over Inverse-Document-Frequency (TF-IDF).
Today's Work : Document Classification Cont'd.
Today's Work : Document Classification Cont'd.
Today's Work : Document Classification Cont'd.
Today's Work : Document(Movies) Classification Cont'd.
Today's Work : Document(Movies) Classification Cont'd. Used Cosine Similarity Matrix to measure Similarity between Top 100 Movies.
Today's Work : Document (Movies) Classification Completed. Used Cosine Similarity Matrix to measure Similarity between Top 100 Movies.
Today's Work : Started new Course 'Sequence Models for Time Series and Natural Language Processing' by Google Cloud (Course 4 of 5 in the Advanced Machine Learning with TensorFlow on Google Cloud Platform Specialization) on Coursera.
Today's Work : Started 'Complete Guide to TensorFlow for Deep Learning with Python' Udemy course.
Today's Work : Introduction to Artificial Neural Network, making ANN from scratch.
Today's Work : Manual Creation of Aritificial Neural Network Cont'd in 'Complete Guide to TensorFlow for Deep Learning with Python'.
Today's Work : Manual Creation of Aritificial Neural Network Completed.
Today's Work : Tensorflow Basics.
Today's Work : Made ANN for Linear Regression using TensorFlow.
Today's Work : Made DNN for Classification problem using TensorFlow.
Today's Work : Deep Learning With TensorFlow Cont'd.
Today's Work : Deep Learning With TensorFlow for MNIST Dataset.
Today's Work : Convolutional Neural Network for MNIST Dataset (Handwritten Digit Recognition).
Today's Work : Convolutional Neural Network for MNIST Dataset (Handwritten Digit Recognition) Completed. Used Google Colab for training the Model. Model built with 99.39% Accuracy.
Today's Work : Introduction to Recurrent Neural Network.
Today's Work : Recurrent Neural Network with TensorFlow Cont'd.
Today's Work : Manual Creation of Recurrent Neural Network with TensorFlow.
Today's Work : Recurrent Neural Network with TensorFlow Cont'd.
Today's Work : Recurrent Neural Network with TensorFlow for Sin(t). Also, learned about Word2Vec Model.
Today's Work : Worked with Abstraction APIs for Tensorflow. Learned about Autoencoders and their use as Dimensionalty Reduction. Setup the environment for 'OpenAI Gym' for Reinforcement Learning.
Today's Work : Bot for CartPole Game using OpenAI Gym with TensorFlow.
Today's Work : Introduction to GANs ( Generative Adversarial Networks ).
Today's Work : GAN for MNIST (Handwritten Digit) Dataset.
Today's Work : Implementation of GAN for MNIST (Handwritten Digit) Dataset.
Today's Work : Deep Learning and Computer Vision.
Today's Work : Face Detection Intuition.
Today's Work : Face Detection using OpenCV
Today's Work : Live Face Detection as well as Face Detection in Images using OpenCV.
Today's Work : Object Detection using SSD (Single Shot Multibox Detection)
Today's Work : Reading of Online Book 'Neural Network and Deep Learning'. Build a Text Summarizer Model by using Natural Language Processing approach.
Today's Work : Reading - Neural Network And Deep Learning.
Today's Work : Reading - Neural Network And Deep Learning, Maths behind Backpropogation.
Today's Work : 'Neurons That Fire Together, Wire Together.' | Backpropogation Calculus | 3Blue1Brown.
Today's Work : Reading | Neural Network And Deep Learning - Backpropogation.
Today's Work : Regression Excercise with TensorFlow.
Today's Work : Unsupervised Deep Embedding for Clustering Analysis (DEC).
Today's Work : Unsupervised Deep Learning.
Today's Work : Basic understanding about Word2Vec Model.
Today's Work : Hands on Word2Vec Model training and testing. Explored Google's Pre-trained Word2Vec Model.
Today's Work : Cluster Analysis using DBSCAN.
Today's Work : Identifying Movies Genres and then Classifying them into their Genre using Cluster Analysis Technique DBSCAN.
Today's Work : Cluster Analysis using Word2Vec and Doc2Vec Model from Gensim. Started new course 'Applied AI with Deep Learning' on coursera.
Today's Work : Environment Setup for IoT. LSTMs understanding.
Today's Work : Re-evaluating Regression Task using Estimator API of TensorFlow.
Today's Work : Style Transfer Using VGG16 through Keras.
Today's Work : Visualizing Data In Higher Dimensions.
Today's Work : 'Titanic: Machine Learning from Disaster' on Kaggle
Today's Work : 'Titanic: Machine Learning from Disaster' on Kaggle Cont'd.
Today's Work : 'Titanic: Machine Learning from Disaster' on Kaggle Completed with 84% Model Accuracy.
Today's Work : 'House Prices: Advanced Regression Techniques' on Kaggle.
Today's Work : 'House Prices: Advanced Regression Techniques' on Kaggle Cont’d.
Today's Work : 'House Prices: Advanced Regression Techniques' on Kaggle Completed.
Today's Work : 'Facial Keypoints Detection' on Kaggle.
Today's Work : 'Facial Keypoints Detection' on Kaggle Cont’d.
Today's Work : Sentiment Analysis of Real Time Tweets Regarding Facebook.
Today's Work : Image Captioning With Keras.
Today's Work : Image Captioning With Keras Cont'd.
Today's Work : Face Aging with Conditional Generative Adversarial Network (GAN).
Today's Work : Face Aging with Conditional Generative Adversarial Network (GAN) | Reading.
Today's Work : Recognizing Hand Written Digits and Characters | Reading.
Today's Work : Recognizing Hand Written Digits and Characters | Reading Cont'd.
Today's Work : Week 3 on 'Machine Learning By Andrew Ng' coursera.
Today's Work : Week 3 on 'Machine Learning By Andrew Ng' coursera. Logistic Regression Model and its Cost Function.
Today's Work : Week 3 on 'Machine Learning By Andrew Ng' coursera. Logistic Regression Model, Simplified Cost Function for Logistic Regression, Advanced Optimization Algorithms and Multiclass Classification.
Today's Work : Week 3 on 'Machine Learning By Andrew Ng' coursera. Regularization to solve the problem of Overfitting. Learned about Regularized Logistic Regression and Regularized Linear Regression.
Today's Work : Week 6 on 'Machine Learning By Andrew Ng' coursera. Evaluating a Learning Algorithm.