mahesh3394 Goto Github PK

bank_loan_taking_prediction

basic-analysis-data-science

Analysis is basic and most important step in any data science project. Few are steps are mandatory and need to follow in each project. Some of them are mentioned here.

basic-webscrapping

This repo is all about basic web scrapping technique for simple use. Web scrapping done using python libraries. This is very useful tool to extract small set of data.

basics-of-lstm

LSTM is long short term memory which is advancement in RNN for long term dependency. For email response generation LSTM is used with pre defined layer and word embedding techniques.

claim-description-classification

This problem is NLP problem of classification type. Where we have to predict certain class using claim description given. There are two class label which are highly imbalanced and need to be handled.

claim-severity-insurance

In insurance domain it is very important to do correct reserving. Hence it is very important to find severity of claims at real time basis. So by using different data features we are finding claim severity. We build simple regression model for same.

claims-feedback-analysis

In insurance domain retention of customer is very important. Thus feedback mail sent to customers. In order to find retention rate it is very important to do the sentimental analysis on feedbacks. Here we are using multiple NLP libraries for doing the sentimental analysis and building a classification model.

clinical_text_classification

Text classification with fine tuned LLM model. Bert model fine tuned on created dataset.

conditional_gan_sketch_color_problem

We are creating a deep learning generative model using conditional gan, which can create a color images using sketch images provided.

email-advertising-ctr-predictions_a

In advertising industry it is very useful to find impact of advertising campaign. While sending Emails is one of the campaign. To see the impact they need to find CTR that click through rate for emails. Here we are predicting a CTR.

mahesh3394

Config files for my GitHub profile.

nlp-attention-english-to-hindi-translator

Attention base Model to translate english text to hindi text. Attention models, also called attention mechanisms, are deep learning techniques used to provide an additional focus on a specific component. In deep learning, attention relates to focus on something in particular and note its specific importance.

popular-graphs-for-beginner

In this notebook we will see the most commonly use Plots/Graphs in data science with Toy case and practical example.

query_response_generation_llama

In this project we hosted LLAMA model with 7B parameter for response generation. Here we created a rest api which can generate a response when provided a query text.

sentimental-analysis-rotten_tomatoes-review-analysis.

Rotten tomatoes dataset consist of movie reviews. By using the we have to do sentimental analysis on reviews and find positive and negative reviews. This is NLP problem where multiple NLP libraries are used to decoding of textual data.

text-classification-using-cnn

Text classification using Convolutional Neural Networks (CNNs) is a popular deep learning technique for natural language processing (NLP) tasks. CNNs use filters to extract features from the text, and then use these features to classify the text into predefined categories.

training-of-transformer-on-dummy-data

Here we try to understand how transformer works and try to replicate architecture from paper published. Also we will train simple architecture on dummy dataset.

understanding-how-optimization-works-sgd

Optimization is very important concept in machine learning or deep learning. We are using multple optimization technique . One of such algorithm is SGC that Stochastic Gradient Descent. There are inbuilt libraries in python for same but in order to understand concept we are creating a code from scrtatch.

understanding-of-backpropagation-and-optimizers-in-deep-learning

Backpropagation is a widely used algorithm for training deep neural networks. It is a method for computing the gradients of the loss function with respect to the weights and biases of the neural network. The gradients are then used to update the weights and biases in the opposite direction of the gradient, with the aim of minimizing the loss .

understanding-of-callbacks

Callbacks are functions used in deep learning frameworks to monitor and modify the behavior of a neural network during training. They are called at specific points in the training process and can be used to perform tasks such as logging metrics, saving model checkpoints, adjusting the learning rate, and stopping training early if a condition met.

understanding-of-hyperparameter-tuning

Hyperparameter tuning is the process of finding the optimal hyperparameters for a machine learning model. Hyperparameters are values that are set prior to training a model and affect its performance, but cannot be learned from the data. Some common examples of hyperparameters include the learning rate, regularization strength.

understanding-of-llm-encoder-and-decoder-model-

Text generation is one of the use case where LLM is used widely. Here text generation problem solved using encoder and decoder model with defining blocks from scratch.

understanding-of-performance-metrics-in-machine-learning

Performance metrics are essential tools in evaluating the effectiveness and efficiency of machine learning models. These metrics are used to quantify the accuracy, precision, recall, and other relevant statistics of a model's predictions.Performance metrics are critical in selecting the appropriate model and evaluating its performance.

understanding-of-vectorizer

A vectorizer is a tool used in machine learning to convert textual or categorical data into numerical vectors that can be understood and processed by algorithms. The process of vectorization involves transforming raw data into a structured format that can be used for modeling and analysis.