Token: TMX

Machine learning platform for crypto trading

ICO dates
Start date: 2018-04-02
End date: 2018-05-02

Registrated in: UK

Platform: Ethereum
Type: ERC20

PREMIUM ICO Screenshot
Trademindx categories
Artificial Intelligence Big Data Cryptocurrency Investment Platform Software
Trademindx token sale
KYC passing required No | Whitelist No | Restriction for countries USA, Singapore, China
Hard cap 2500 ETH
Tokens for sale 500,000,000
Token distribution in ICO
Price in PreICO 1 TMX = 0.000005ETH
Minimal investment 1000.00 USD
Acceppting ETH
Bonus in ICO 20% referral program Pre-sale bonus $1k-10k 20% Pre-sale bonus $10k-15k 30% Pre-sale bonus $50k+ 50%
Trademindx news, social
Trademindx search trends in Google
Random whitepaper excerpts

November 2017
(later simply referred to as “the system”) aims to
facilitate trading decisions through applied natural-language processing,
sentiment analysis and machine learning algorithms. Built as a micro-ser-
vices based distributed architecture, its primary function is to analyse large
volumes of data and provide clear trading indicators to the client in real-
2 of 17

1. Introduction
November 2017
The system combines natural-language processing, sentiment analysis and ma-
chine learning algorithms to classify large volumes of data into clear (BUY, SELL or
HOLD) trading signals for the trader in real-time. The default approach is implemented
using a sentiment analyzer plus a maximum entropy (MaxEnt) algorithm with
The former quantifies sentiment. The latter applies a multinomial logistic regression
model [1] to categorise data and continually improve accuracy.
One of the design features of the system is that there is no requirement to train
the model. In this respect, it differs quite significantly from other machine learning im-
plementations. The system uses its own unique algorithms to
train itself
, generating
models, checking results against known data and subsequently refining those models.
The system is by design, in a state of continuous automated machine learning, constant-
ly adjusting and upd...

November 2017
Moreover, the system aims to provide more granular outputs which are not nec-
essarily binary (positive/negative). The system can provide more complex outcomes
which may include additional characteristics such as time horizons — e.g.
buy and hold
for 1 month
. However for simplicity, the initial implementation provides buy, sell or hold
outputs with a fixed 24hr time horizon.
4 of 17

2. High-level Algorithm Overview
November 2017
• Opinion moves the market.
• Can Artificial Intelligence sense opinion?
• And, can AI predict the market before it moves?
What impacts the price of precious metals, currencies, cryptocurrencies, stocks?
Some may think it’s the money invested, but this is not the case. It is driven by opinion.
To be more precise, it is opinion and consensus.
Since it is generally accepted, let’s just take it is as a baseline assumption: opin-
ion moves the market.
What if you could model the set of traders participating in a particular market and
deduce their overall consensus on opinion? What do you get? A super trader. That trad-
er can clearly predict the market movements correctly. Now it’s just a case of building
such a trader with AI technology.
There’s an obvious problem, not all traders are actively commenting on social
media or forums! However, we can assume that we are taking a slice of the market and

3. Applying Artificial Intelligence
November 2017
AI can be used to measure opinion to a statistical degree of confidence.
In order to attempt to measure opinion, we need to source raw information from a
number of sources, apply Natural Language Processing (NLP) to make sense of it, turn-
ing it into data that can be processed by a machine. Then, we apply sentiment analysis
plus machine learning algorithms to derive trading signals. Raw information sources in-
• News articles
• Twitter feeds
• Reddit posts
• Other social media feeds
• Forum threads
• Even GitHub repository activity
• And of course, market data feeds from cryptocurrency exchanges
6 of 17

November 2017
Once we get the raw data, we run it through a sentiment analyzer which allocates
a sentiment measure. For example:
SentimentConfidence {
In the case above, the system has deduced that at that specific moment in time,
after analyzing 781 samples of discreet data, Ethereum has an overall positive senti-

with a 57% degree of confidence. We then repeat this for all other cryptocurren-
cies that we want to analyze. We then track sentiment as it changes in real-time.
Research has shown that sentiment analysis algorithms are approaching high
degrees of accuracy. Our sentiment analyzer is based on VADER

Valence Aware Dic-
tionary and sEntiment Reasoner, which has been shown to p...

November 2017
time feeds, which is in direct contrast to mature forex and stock markets. So our market
data sources are abundant.
Furthermore, what makes the cryptocurrency market such an interesting domain
to analyze, is the relative immaturity which leads to high volatility.
Market moves are often triggered by announcements, news articles, support by
social influencers and even rumours. This often happens very quickly, so if an AI algo-
rithm could sense changing opinion, then this could in turn provide a trading advantage.
And of course, a trading advantage has intrinsic monetary value

As well as taking sentiment data, we also take our input data and attempt to build
AI models and then refine those models using feedback. This is where machine learning
comes in. We back test our models against known market moves and then continuously
re-calibrate those models, thereby attempting to improve accuracy. The model starts off

November 2017
For our models, we have decided to use Maximum Entropy (MaxEnt) modelling
which has been shown to produce more accurate results than Naive Bayes. However, in
the future models can be swapped out if others are deemed to be more accurate.
9 of 17

Trademindx Roadmap

Project Launch - Q4 / 2017

Trademindx project launched / whitepaper published / prototype development begins / build out of core services.
Prototype Development - Q1 / 2018
Complete prototype / architecture and tech stack design / roadmap and faq published / smart contract development in Solidity / token pre-sale.
Token Sale - Q2 / 2018
TMX token sale / smart contract on Ethereum blockchain (ERC20) / compatible with Ethereum wallets / exchange listing / airdrops.
Artificial Intelligence - Q3 / 2018
AI model training and calibration for all cryptocurrency coins/tokens with a market cap of more than $100mm.
Token Burn - Q4 / 2018
Unallocated TMX tokens will be burned off (destroyed) thus reducing circulating supply.
Beta Release - Q1 / 2019
Trademindx closed beta for TMX token holders / feedback from community / bug fixes and improvements.
GO LIVE! - Q2 / 2019
Trademindx platform goes live / subscription based pricing model using TMX tokens!
Enterprise Edition: TaaS - Q4 / 2019
Trademindx as a service / onboarding of enterprise clients / custom strategies and algorithms / strategic partnerships.
Hedge Fund Setup - Q2 / 2020
Automated algo hedge fund driven by Trademindx algorithms with a range of products to suit cryptocurrency investors with various risk appetites / TMX token utilisation.
Denis Golovenko
Denis Golovenko Financial Advisor

Raghu Rughani
Raghu Rughani Technical Advisor

Saeed Assou
Saeed Assou Community Manager and Investor