pec-css / slowmokit Goto Github PK

Library which can be used for machine learning in C++.

Home Page: https://pec-css.github.io/slowmokit/

License: GNU General Public License v2.0

C++ 23.72% CMake 36.82% Rust 29.54% Shell 9.93%

cpp machine-learning ml-library classification-algorithm clustering cplusplus cplusplus-20 machine-learning-algorithms regression-models slowmokit

slowmokit's People

Contributors

Stargazers

Watchers

Forkers

ken1000minus7 divyansh1011 ishwarendra uttammittal02 sakshambeniwal sugam-arora armaanbadhan yashitabansal sar-thak-3 modernbeast02 nandinigera abhinavmalhotra01 daksh0786 harshjohar sparsh44 dhruv-porwal

slowmokit's Issues

N-matrix declaration

Declare a N dimensional matrix class.

constructors

Implement vectorization in c++

Refer to this Research paper for reference

Feature: Implement Ridge and Lasso regression

Implement mini-batch gradient descent

Feature : Implement Random Forest as a classifier

Feature: Implement Gaussian Naive Bayes

KNN classifier

KNN classifier in c++

Feature: Implement train test split for pre-processing

Implement Silhouette Score

Feature: Implement SVM for Classification

Fix CMake to make all the examples running

Feature: Implement DBSCAN clustering algorithm

Feature: Implement csv reader

Feature: Implement matrix multiplication

Feature: Logistic regression model

Implement a logistic regression model

Feature: Implement various distance metrics

Euclidean, Manhattan, Minkowski, cosine similarity

Features: Implement addition, subtraction, of two matrices.

Easy Category.

Feature: Implement the code to plot a dendogram

The dendogram should give different colours to different clusters based on the Agglomerative Clustering as the below image shows that there should be 2 clusters in the dataset

Feature : Implement Decision Tree for regression

Implement standardization for data pre-processing.

Implement confusion matrix (with heatmap) for multi-class classifier

Metrics: Implement F1-score for multi-class classification

Overloading [] and << operator

Easy Category
Accessing matrix like vector and printing it as we need will make it easier for people to use it.

Implement a classification report for a classifier

To see what all things are required in the report, you can refer here

Implement addition and subtraction of matrix with a scalar

Overload operator+ operator- and operator *. Function signatures are present in matrix.h already at Line-145, 146, 155 and 156.
Also overload operator += and operator -= for a matrix and scalar (Function signature for this is not present so add that in matrix.h too.)

Feature: Implement Radial Basis Function (RBF) for SVM

Implement accuracy

Feature : Implement Random Forest for regression

Feature: Implement SVM for Regression

Implement precision and recall for a multiclass classifier

Feature: Implement Polynomial Kernel for SVM

Automate the addition of hpp files to avoid conflicts

K means clustering

Feature : Implement Agglomerative Clustering

Decision Tree model

Ml decision tree model

Feature: Implement normalization for pre-processing

Automate the creation of all files needed to add a model

Feature: Implement one hot encoder for pre-processing

Cleanup metrics directory

Create a metrics class in metrics dir and add all 5 functions as static functions.
Test all function's working before making a PR

Implement Label Encoder for preprocessing

Feature : Implement Categorical Naive Bayes

Feature: Implement Gaussian Kernel for SVM

Feature: Implement Multinomial Naive Bayes

Feature: Implement Bernoulli Naive Bayes

Feature : Implement Complement Naive Bayes

Make a cpp lint tool to check the cases and spaces

Feature : Implement a basic neural network architecture

Implement Mean-Squared Error

Implement matmul, subtract, add function.

Add function signature for each function inside the Matrix class in matrix.hpp and also implement in matrix.cpp.
Function specifications:
- matmul: Multiplies two Matrices. A Matrix and a scalar is not allowed. Also throw appropriate exception if matrix dimension are not compatible for multiplication.
- add: This function will have two overloads one which will add matrix and scalar and the other one for two matrices. Throw appropriate exceptions if matrix have incompatible dimensions.
- subtract: Everything same as add function but you should do subtraction instead of addition.
Don't implement from scratch, use the already implemented functions operator +=, operator -= and operator *=. This way you will not have to worry about throwing appropriate exceptions as the already implemented function will do it for you.
Don't forget to add comments above each function signature you add in matrix.hpp. Avoid adding comments in matrix.cpp unless very important.

Feature : Implement Various Activation Functions

Sigmoid, ReLU, tanh, tan-1, softmax, binary step, Leaky ReLU