im-pek Goto Github PK

followers: 18.0 following: 31.0 repos: 41.0 gists: 0.0

Name: Pek Yun Ning

Type: User

Bio: Data Scientist | with exceptional focus on NLP, Text Analytics, Topic Modelling, Recommendation Systems, Trend Analysis, Pattern Recognition, Novel Algorithms

Location: Singapore

Blog: https://www.linkedin.com/in/pek-yunning/

Pek Yun Ning's Projects

add-symbol-currency-2dp-to-every-excel-cell

Add symbol, currency, and/or 2 d.p. (to every excel cell)

amazon-textract-enhancer

This workshop demonstrates how to build a Document parser and query engine with Amazon Textract and other services, such as ElasticSearch and DynamoDB.

autofit-excel-cell-widths-using-xlwings

Autofit excel cell widths (using xlwings). Xlwings is an extremely efficient, state-of-the-art python library to manage, edit, and manipulate excel files & data. Its documentation can be viewed at https://readthedocs.org/projects/xlwings/downloads/pdf/stable.

bold-excel-cells-using-openpyxl

Bold Excel Cells (using openpyxl)

classify_nouns-verbs-adjectives

This is how nouns, verbs, and adjectives can be classified from a bunch of text.

combine-excel-sheets-using-openpyxl

Combine excel sheets from two workbooks (using openpyxl)

corex_topic

Hierarchical unsupervised and semi-supervised topic models for sparse count data with CorEx

corex_topic_modelling

Correlation Explanation (Corex) is a topic modelling technique that is great at identifying 'hidden' topics, or low-frequency-worded but representative topics, very well. It was originally created by Greg Ver Steeg.

csv-blank-removal

Removes blank cells in CSV files using Python. In Python list, it is seen as 'nan'.

data-anonymisation

Anonymising data in meeting notes.

doc2vec_topic_modelling

Doc2vec method of topic modelling. It's document-level of Word2vec. Builds on the concept of word vector representations.

doc2x_topic_modelling

Doc2X is a novel topic modelling technique created in June 2019, by yours truly, Pek Yun Ning. It hybridises the older Doc2Vec and Corex topic modelling algorithms to form this all-new algorithm.

docs

TensorFlow documentation

enumerate_using_python

A simple implementation of 'enumeration' in Python. In this case, we number webchats from one whole chunk of text filled with tons of webchat entries.

excel--unique-values-in-columns-and-difference-between-two-columns

Excel - find unique values in columns, & difference between two columns

fasttext_topic_modelling

FastText is a topic modelling technique originally created by Facebook AI Research team. Its first stable version release was on December 2018. FastText is now available on Python's gensim and scikit-learn as well.

geographical_hexbins-projections

Data science applications in geographical data. Involves hexbins and projections.

lda2char_topic_modelling

LDA2Char is a novel topic modelling technique created in June 2019, by yours truly, Pek Yun Ning. It hybridises the older LDA and FastText topic modelling algorithms to form this all-new algorithm.

lda2word_topic_modelling

LDA2Word is a novel topic modelling technique created in June 2019, by yours truly, Pek Yun Ning. It hybridises the older LDA and Word2Vec topic modelling algorithms to form this all-new algorithm.

lda2x_topic_modelling

LDA2X is a novel topic modelling technique created in June 2019, by yours truly, Pek Yun Ning. It hybridises the older LDA and Corex topic modelling algorithms to form this all-new algorithm.

lda2xpand_topic_modelling

LDA2XPand is a novel topic modelling technique created in June 2019, by yours truly, Pek Yun Ning. It hybridises the older LDA, Corex, and Word2Vec topic modelling algorithms to form this all-new algorithm.

lda_topic_modelling

Latent Dirichlet Allocation (LDA) is a topic modelling technique that involves a three-layered probabilistic approach, taking into account words at words, documents, and corpus level. It is accompanied by its very own unique and powerful data visualisation tool, LDAvis (as part of this code in its Pythonic version, pyLDAvis), as well.

logistic_regression_-_confusion_matrix

An application of Logistic Regression and presentation of its accuracy and precision using Confusion Matrix. This was applied to a consumer complaints dataset.

model_augmentation

Model Augmentation by adding more sample data points to an existing (small) dataset in a mathematical manner. Carried out using mathematical concepts such as Euclidean Distance and Uniform Criterion (max-of-min concept). This adds exploration factor to the model, reduces prediction error, and improves global accuracy (in optimisation).

im-pek Goto Github PK

Pek Yun Ning's Projects

Recommend Projects

Recommend Topics

Recommend Org

Jobs