Thomas T. Profile

3,8

(2 opinii)

Zacząłeś obserwować

Błąd związany z obserwowaniem użytkownika.

Ten użytkownik nie pozwala użytkownikom na obserwowanie go.

Już obserwujesz tego użytkownika.

Twój abonament pozwala na obserwowanie 0 projektów. Zmień poziom abonamentu na wyższy tutaj.

Udało się przestać obserwować

Błąd w zakończeniu obserwowania użytkownika.

Udało Ci się polecić

Błąd podczas polecania użytkownika

Coś poszło nie tak. Proszę odświeżyć stronę i spróbować ponownie.

Udało się zweryfikować adres e-mail.

$30 USD / godz.

singapore,

singapore

$30 USD / godz.

Obecnie jest 3:22 PM tutaj

Dołączył lipca 11, 2018

0 Zalecenia

Thomas T.

@ThomThio

3,8

(2 opinii)

3,9

100%

$30 USD / godz.

singapore,

singapore

$30 USD / godz.

80%

Ukończone Projekty

100%

W Budżecie

38%

W czasie

N/A

Powtórna Stawka Zatrudnienia

Machine Learner, Python Wrangler, Java Drinker

Selected portfolio: [login to view URL] Familiar with various languages (C#, C++, Android (Kotlin and Java), JS, Python) and application of supervised learning on the ML front. 3 years in product management of a custom enterprise software, so I know how to take in and integrate complex requirements, and organize your needs to get the job done I primarily focus on data prediction and analysis, back-end development and simulation/modelling projects.

Wyświetl Wszystkie eCommerce Developers w Singapore

Skontaktuj się z Thomas T. w sprawie projektu

Zaloguj się, aby omówić szczegóły na czacie.

Portfolio

Analysis @ https://drive.google.com/file/d/14hWUaUjdbZB_aToj6IJE7P12B92HOLwy/view

Repo @ https://github.com/ThomThio/SurgePricingSimulation

Using NetLogo software, this is a Multi-agent simulation of viable strategies drivers and taxi companies can take in response to private hire (Uber/GrabTaxi) surge pricing. We built price sensitivity, surge pricing models and demand allocation by regional locations using data from Fujitsu-SMU Urban Computing & Engineering (UNiCEN) Laboratory.

A user would start the simulation where it populates taxis across Singapore following the distribution of data by UNiCEN, thereafter following paths of each data record from point to point in an area.

Thereafter, each taxi type (private hire or normal) compete for customers, and P&L is logged as an aggregate for the company. These are displayed in real-time graphs.

Additionally, users can opt to 'play' as a taxi driver, choosing a range of pre-set strategies to try to outcompete the private hire m

Decision support via Uber/Grab vs Taxis Simulation

Client wanted to his robot to recognize specific items, picking them up or avoiding them.

Used tensorflow's neural networks. Trained Object recognition model to detect custom items and worked with other developers to customize actual robot movement for collision detection together with object recognition.

Client is able to build a robot to deliver or transport goods, and have ability to train to any other object.

Object recognition on Raspberry Pi Robot

Analysis: https://docs.google.com/presentation/d/1rJcNyATxGjhqJxAUAqLiu04Gl5yMlKFDQ1US2HwZwNU/edit#slide=id.p

Coursework for Excel Modelling: Seasonality analysis of seafood inventory, best client-location route optimizer, order dispatcher, inventory tracker and re-stocking level prompts.

Inventory Planning, Scheduling and Prediction of Seafood

4289921

Data science: Predicting Store Sales, top 4% performance

4289918

Web scraping and data import

Analysis: https://wiki.smu.edu.sg/1617t1IS428g1/IS428_2016-17_Term1_Assign1_Thomas_Joseph_Thio_Kit_Sun

Dive into Singapore's private property market in 2015 to discover trends, insights and possible explanations of phenomena. Done as part of module in SMU's Visual Analytics for Business Intelligence

Real estate analysis: Demand and Supply in SG (2015)

Decision support via Uber/Grab vs Taxis Simulation

Object recognition on Raspberry Pi Robot

Inventory Planning, Scheduling and Prediction of Seafood

4289921

Data science: Predicting Store Sales, top 4% performance

4289918

Web scraping and data import

Real estate analysis: Demand and Supply in SG (2015)

Oceny

Zmiany zapisane

Pokazuje 1 - 2 z 2 komentarzy

Filtruj oceny po przez:

5,0

$65,00 USD

Python Code to get keywords as input, analyze them in google search results and provide level of confidence

Very knowledgeable. I would definitely recommend.

Python

Web Scraping

Machine Learning (ML)

Data Mining

Farid M.

@faridbayan88

•

5 lat temu

3,0

$105,00 USD

custom thinkorswim strategy

he was sure it can be don , but when he actually started it , he found out that he can't do it .

PHP

C Programming

Software Architecture

Metatrader

C++ Programming

Fisal S.

@fs4590

•

5 lat temu

Doświadczenie

Proprietary Trading

lis 2016 - Obecnie

• Built redundancy and risk systems working with IB TWS, MultiTradex & Wenhua. Used C#, C, Python, Javascript • Sped up position reports from 40 to 5 seconds by optimizing database queries and warehouses in MSSQL and SQL Server • Built real-time spread price visualization and GUI for changing Lookback and Interval parameters using Python, Javascript • Researching and developing strategies on Quantopian & QuantConnect.

Boutique Management Consulting Firm

gru 2017 - lut 2018 (2 miesiące, 1 )

• Research on Fama-French factor models and discovery of alpha and beta with alternative data sets. • Built knowledge graph of documents, allowing searching of key topics via semantic relatedness and authority of papers using graph-tools and Django.

SAS

kwi 2017 - cze 2017 (2 miesiące, 1 )

• Designed and built API controller for developers and pre-sales teams to build, call and connect sales demos and other models to their front-end platforms. Slashed production time by 30%. • Prepared sample demos in Amazon Alexa & SAS VIYA integration and Web Chatbots, using SAS Viya, VDDML, Python, Javascript, Linux, Motion.ai, Amazon Alexa and Lambda

Edukacja

Bsc. in Information Systems

Singapore Management University,

Singapore

2014 - 2017

(3 lata)

Kwalifikacje

Coursera

2014

The course is a hands-on application of machine learning concepts to kick start the understanding behind the process and models, supervised and unsupervised.