Intelligence Can Solve Complexity

Develop Your Machine Learning Trading System

Starter Notebook Join Community
Tradeset
ML-Ready Features API

ML-Ready Features

We discovered that the key to bridging the gap between data scientists and quantitative trading lies in ML-ready features. Crafting insightful features from financial data demands expertise in finance and trading, skills not universally held by data scientists. We offer proven ML-ready features tailored for each asset, enabling data scientists to leverage their skills for accurate predictions. Access historical ML-ready features for free. Note: Unlike text, images, and some tabular data, deep learning is not effective for extracting and selecting features from raw financial market data.

python 
	from tradeset import get_features 
	get_features(asset_name, api_key=API_KEY, feature_type='train') 
	df_features = pd.read_parquet(f"{asset_name}_train.parquet")

Target Definition

Defining targets (labels) is an imperative step in developing a profitable prediction model. You can define classification target parameters and we will provide labled data for you. Explore a sample in our Starter Notebook. You can access the target definiton service for free.

python 
	from tradeset import create_target 
	asset_name = 'USDJPY' #Define the forex pair 
	trade_mode = 'long' #(long or short) 
	target_look_ahead = 315 #look-ahead period in minutes. 
	target_take_profit = 30 #profit in pips 
	target_stop_loss = 6 #stop loss in pips 
	target_token, target_name = create_target(
	   asset_name,
	   trade_mode,
	   target_look_ahead,
	   target_take_profit,
	   ftarget_stop_loss,
	   API_KEY,
	)
	df_target = pd.read_parquet(f"./{target_name}.parquet")

Strategy Definition & Backtest

Prediction is just one aspect of successful trading. Formulating strategies and backtesting are equally crucial. Our flexible framework allows you to optimize parameters, potentially turning weak predictions into profitable signals. Explore sample strategies in our Starter Notebook. You can access these components via API services, all FREE for historical experiments.

python 
	from tradeset import backtest_strategy 
	strategy_config = {
	   'target_token': target_token,
	   'volume': 0.1,
	   'initial_balance': 3000,
	   'spread': 2,
	}
	backtest_results, backtest_df = backtest_strategy(
	   df_model_signal[["model_prediction"]],
	   strategy_config,
	   API_KEY,
	)

Keep Your Experiments Organized

With our platform, you can track and manage your submitted backtests, leveraging our comprehensive suite of tools to compare and analyze various combinations of models, targets and strategies, ultimately identifying the optimal approach for your trading system.

dashboard

Real-time Services

After creating profitable models and strategies through free historical experiments, you can request real-time services including real-time features, execution expert bot and cloud services for Demo and Live trading.

python 
	Comming Soon

About Tradeset

Forecasting and trading in financial markets require a blend of human intelligence and artificial intelligence (AI). At TradeSet, we are committed to democratizing algorithmic trading for all data scientists. By providing a platform equipped with proven ML-ready features, we empower data scientists to concentrate on crafting ML models, defining labels, and devising strategies. Our services remove the barrier of needing extensive financial market expertise, streamlining the process of building profitable quantitative trading systems.