yanliang1102 Goto Github PK

2014

Official content for the Fall 2014 Harvard CS109 Data Science course

2014_data

Data directory for the CS109 Data Science course

advanceddb_code

angular-and-aspnet-build2014

Source code from the Build 2014 presentation by David Catuhe and Jon Galloway

angular-lynda

Building a Data-Driven App with AngularJS with Ray Villalobos

autobuildservertesting

blaze-tutorial

compiler_project_ou

computer_graphics_ou

content

Official content for Harvard CS109

course-neural-networks-for-machine-learning

Neural Networks for Machine Learning (University of Toronto)

coursera-natural-language-processing-specialization

Programming assignments from all courses in the Coursera Natural Language Processing Specialization offered by deeplearning.ai.

cryptography-code

cs224n

CS224n: Natural Language Processing with Deep Learning Assignments Winter, 2017

deeplearning

dmlc-mxnet

easynlp

EasyNLP: A Comprehensive and Easy-to-use NLP Toolkit

eventextractionpapers

A list of NLP resources focused on event extraction task

free-programming-books-zh_cn

免费的计算机编程类中文书籍，欢迎投稿

hmeae

Source code for EMNLP-IJCNLP 2019 paper "HMEAE: Hierarchical Modular Event Argument Extraction".

jave

leetcode

lihang-code

《统计学习方法》的代码实现

machine-learning-classify-handwritten-digit

Classify handwritten digits using machine learning techniques Yan Liang, Yunzhi Wang and Delong Zhao Project scope For our machine learning project, we propose to build several machine learning classifiers that recognize handwritten digits. Handwritten digit recognition is a classic problem in machine learning studies for many years. We plan to do several experiments using different machine learning algorithms and compare the pattern recognition performance. We hope to create a classifier that has same or better categorization accuracy than record performance from previous studies. Yan will focus on neural network, Delong will focus on the random forests methods, and Yunzhi will focus on SVMs and KNNs. We will also develop a final novel classifier that combines the best models from our different experiments. We hypothesize that the final classifier will archive a categorization accuracy of 0.99. This indicates that the classifier correctly classified all the handwritten digits but 1% of the images. The goal of handwritten digit recognition is to determine what digit is from an image of a single handwritten digit. It can be used to test pattern recognition theories and machine learning algorithms. Preprocessed standard handwritten digit image database has been developed to compare different digit recognizers. In our semester project, we will use modified National Institute of Standards and Technology (MNIST) handwritten digit images dataset from kaggle digit recognizer project. The Kaggle MNIST dataset is freely available and collected 28,000 training images and 42,000 test images. Each image is a preprocessed single black and white digit image with 28 x 28 pixels. Each pixel is an integer value range from 0 to 255 which represent the brightness of the pixel, the higher value meaning darker. Each image also has a label which is the correct digit for the handwritten image. For each input handwritten image, our model will output which digit we predict and evaluate with the correct label. We will use 28,000 training images to train our machine learning model and use 42,000 test images to test the performance. Then we will calculate the percentage of the test images that are correctly classified and compare the performance of different machine learning algorithms.

maven-dataset

Source code and dataset for EMNLP 2020 paper "MAVEN: A Massive General Domain Event Detection Dataset".

ml_spark

machine learning code in spark

n3-collection

N3 - A Collection of Datasets for Named Entity Recognition and Disambiguation in the NLP Interchange Format

ok-coders-spring-2015

ou-ise-stock-prediction

OU-ISE5103-FinalProject-StockPrediction

parallel-programming-yan

parallel programming code