bnicholl Goto Github PK

3_class_svm.py

utilize support vector machine algorithm in order to classify three species from the iris dataset based on their sepal length and their sepal width

astar_pathfinder

traverse through any binary array using the Manhattan distance by giving a start node location and end node location

bnicholl.github.io

boomi_ws

Boomi WS data

churn_for_workshop

cluster_every_combo

Runs through a database of features, and clusters every single combination. amount_of_combinations parameter gives you the option to tell the algorithm how many features you want to cluster at a time.

connect_four_intelligent_player

Intelligent program that utilizes a depth three MinMax algorithm. A five dimensional array of arrays was used in order to hold all of of the possible states.

contractorapp

The bees knees

crypto_currencies_analyses

analyses of 4 crypto currencies

deep-cnn-for-image-recognition

neural network that learns to classify images

exo_data

gradientdescent_vs_newtonoptimization

Differences, strengths and weaknesses of two completely different optimization algorithms

iterate_through_infinity

Iterate through a list of nodes where each dimension could theoretically be infinity. Here we take a look at utilizing recursive algorithms using DFS, and iterative algorithms using BFS.

loan_risk

kaggle competition: trying to predict the probability pertaining to someone defaulting on a loan given a database of features

melting_temp_dna

Gives you the melting temperature of DNA sequences based off of nucleotides. DNA sequences can be uploaded via Entrez from NCBI's website

one_hots_with_window

creates one hot vectors with corresponding labels based on window size

physics

plotly_ipywidget_notebooks

probability_from_postgres

Kaggle competition where the goal is to create an algorithm that detects the probability that an app was downloaded given an add was clicked

probability_word_jawnski

given a string of words, calculate the probability of the topic

redux

Predictable state container for JavaScript apps

rolling_average_solution

star_type_csv

torchnn

basic torch feed forward neural net

transcription_algos

transformer_multi_label

There are no transform neural networks that are able to perform multi label classification tasks. So I engineered one myself

word2vec_neural_net

Word2vector neural network that uses skip-gram model to create word vectors from binary one hot encoded vectors

word_dimensions

Jupiter Notebook describing how words derive meaning for NLP products such as Alexa and Siri