Mando

​Mando Iwanaga
206-747-2291 | ​mandoiwanaga08@gmail.com​ | ​Github​ | ​Blog​ | ​LinkedIn

Data scientist and machine learning engineer with a passion for problem solving, optimization, and efficiency. Experience in data acquisition and data modeling, statistical analysis, machine learning, deep learning, and NLP. With a background in Customer Service, I bring strong skills in team-work, communication, and business understanding that will add value to your industry.

TECHNICAL SKILLS

CORE TOOLS

Python, Pandas, Numpy, Selenium, Tensorflow, StatsModels, OOP, APIs

MACHINE LEARNING

Sci-kit-Learn, Gensim, Keras

DATA CONTROL + STORAGE

Github, Git, PostgreSQL, MongoDB

VISUALIZATION DEPLOYMENT

Matplotlib, Seaborn

OTHER

Flask, HTML/CSS, AWS EC2 Jupyter Notebooks and Lab

TECHNICAL PROJECTS

DreamJobber -​ ​Github​ | ​Live App
A cold-start recommendation system for tech jobs based on a user’s preferences.

●  Acquired job data via web-scrape using Selenium and stored as JSON files
●  Conducted Natural Language Processing using NLTK and Gensim to tokenize 20,000 job descriptions
●  Implemented LDA Topic Modeling to group jobs by topics
●  Applied an Unsupervised Nearest Neighbors model using Euclidean Distance to make recommendations
●  Deployed as Flask webapp where users can access at ​www.dreamjobber.online

Board Game Rank Classifier – ​Github
A classification model to determine if a board game would rank top 25% on BoardGameGeek.com

●  Utilized a mix of API calls and web-scraping to derive data from BoardGameGeek.com
●  Applied Gradient Boosting as final classification model evaluated by ROC score
●  Performed hyperparameter tuning to improve ROC score and Cross Validated model

Football Analytics – ​Github
Football (Soccer) Hypothesis Testing

●  Queried and extracted data from postgreSQL database
●  Constructed null-hypotheses and hypotheses; performed one-sample t-test and two sample t-test
●  Made conclusions based on test results of whether to reject or accept the null hypotheses