Appendix
Glossary
The following is a glossary of terms related to the Parrot project:
Quantitative trading: The process of using mathematical models and computer algorithms to execute transactions to improve the accuracy and efficiency of trading decisions.
Intelligent trading: The process of using artificial intelligence technology for transaction decision-making and execution to improve the intelligence and automation of transactions.
AI model: A mathematical model built based on artificial intelligence technology, used to analyze market data and predict future trends to support intelligent trading decisions.
Quantitative strategy: a trading strategy based on mathematical models and statistical analysis, used to identify trading opportunities and execute trading operations to achieve stable investment returns.
Trading signals: Trading instructions generated based on market analysis and model predictions, used to guide trading operations and execute buying and selling transactions.
Data preprocessing: Perform operations such as cleaning, denoising, and filling in missing values on the original data to improve data quality and integrity and prepare for model training and analysis.
Feature engineering: The process of feature extraction and feature selection on data to extract effective feature information for model training and analysis.
Model training: The process of training and optimizing a model using historical data to improve the model's predictive and generalization capabilities.
Model evaluation: The process of evaluating and validating a trained model to evaluate the performance and accuracy of the model and provide a reference for model tuning and improvement.
Real-time trading: The process of generating trading signals based on the latest market data and executing trading operations in real time to respond to market changes and fluctuations.
Risk management: The process of evaluating and managing risks in transactions to reduce the volatility and risk of loss of the investment portfolio and protect the safety of investors' funds.
Data security: The process of using encryption technology and security protection measures to ensure the security and confidentiality of transaction data and user information.
User identity verification and authorization: The process of verifying and authorizing user identity to ensure the security and legality of the trading platform and prevent illegal operations and fraud.
Parrot’s AI intelligent training model content (part) :
The training strategy of PT quantitative AI large model includes :
Bayesian Dynamic Optimization Model
This is a dynamic optimization model based on Bayesian statistical theory, which is used to update the model when new data is continuously observed and make optimal decisions based on updated information. The model combines Bayesian inference and optimization methods to solve optimization problems in dynamic environments.
Black-Scholes Model
The Black-Skoll model is a financial mathematical model used to price option contracts. It is based on the theory of stochastic differential equations, which describes the random fluctuations of asset prices, and based on this, calculates the theoretical price of options.
Markov Risk Adjustment Strategy
This is a risk adjustment strategy that utilizes Markov chains to dynamically adjust the allocation of a portfolio to maximize returns and reduce risk based on the transition probability between the current state and historical states.
Monte Carlo Maximization Scheme
The Monte Carlo method is a numerical calculation method based on random sampling that is used to simulate random phenomena and estimate probabilities and benefits through a large number of random experiments. The program is optimized using Monte Carlo methods to maximize investment returns.
Black-Litterman Strategy
The Black-Ratton strategy is an asset allocation strategy based on Markowitz portfolio theory and Bayesian inference that is used to adjust standard portfolio models to reflect investors' subjective views and information, thereby improving portfolio expectations. Yield and risk.
Gaussian-Hermes Model
This is a mathematical model based on Gaussian distribution and Helmers function that is used to describe the stochastic volatility of asset prices and is used in financial risk management and pricing models.
Weber Volatility Control Scheme
This is a volatility control scheme based on the Weber distribution that is used to dynamically adjust asset allocation in a portfolio to control the risk level of the portfolio and maximize returns.
Hermes-Hertz-Frey-Bernoulli Model
This is a model that integrates mathematical theories such as Hermes, Hertz, Frey and Bernoulli to describe the volatility and randomness of asset prices in financial markets and provide investors with more accurate risk assessments. and basis for decision-making.
Poisson-Laplace Adjustment Model
It is a mathematical model based on the Poisson distribution and Laplace transform that is used to adjust and smooth financial time series data to reduce noise and improve forecast accuracy.
Kalman Filter Optimization Scheme
The Kalman filter is a recursive algorithm used to estimate the state of dynamic systems from a series of incomplete or noisy observations. This solution uses Kalman filtering for optimization to improve prediction and decision-making effects on market conditions.
Fermat's Rule Risk Control Scheme: Fermat's Rule is a mathematical method for controlling risk, based on probability theory, by allocating funds in the investment portfolio to minimize potential losses, thereby achieving effective risk management. manage.
Bessel Weighting Adjustment Scheme: This scheme is named after the 18th-century German mathematician Friedrich Bessel. Bessel weighting scheme is a method for adjusting asset weights, especially suitable for portfolio optimization and asset allocation. Through Bessel weight adjustment, the risk and return characteristics of the investment portfolio can be optimized based on the historical volatility and correlation of assets.
Hamilton Dynamic Arbitrage Model: The Hamilton Dynamic Arbitrage Model is a trading strategy based on arbitrage opportunities, which achieves profits by tracking price differences or arbitrage opportunities between different markets. This model uses mathematical methods and algorithms to identify and exploit temporary imbalances in market prices.
Galton-Randall Strategy: This strategy comes from the names of two economists. The Galton and Randall strategy is a market-neutral strategy that seeks to profit from market volatility by trading both long and short positions simultaneously. This strategy relies on in-depth analysis and prediction of market trends and price movements.
Franz-Maxwell Dynamics Scheme: Franz-Maxwell Dynamics Scheme is a trading strategy based on the dynamic principles of physics. It identifies the market by simulating the dynamic changes and behavior of the market. underlying patterns and make trading decisions.
Nash-Williams Optimization Strategy: The Nash-Williams Optimization Strategy is a trading strategy based on optimization theory that aims to maximize the return of a portfolio while controlling risk. This strategy utilizes mathematical models and algorithms to determine optimal trading decisions.
Taylor-Feyrie Risk Adjustment Scheme: The Taylor-Feyrie Risk Adjustment Scheme is a method used to adjust the risk of an investment portfolio by evaluating and adjusting the risk characteristics of the assets to achieve the optimal performance of the investment portfolio. Risk control and optimization.
Mean Reversion: Mean Reversion theory believes that price fluctuations will return to their long-term mean. Based on this theory, traders will try to trade when prices deviate too far from the mean in the hope that prices will return to the mean level.
Trend Following: Trend following theory believes that price trends have inertia, that is, the current price trend is likely to continue. Based on this theory, traders will try to trade in the direction of the trend in the hope of making profits from the continuation of the trend.
Swing Trading: Swing trading is a trading strategy that takes advantage of short-term fluctuations in market prices. Traders will try to buy and sell in the bands of price fluctuations to make profits between bands.
Contrarian Trading: Contrarian trading theory believes that there are a lot of irrational behaviors in the market. Therefore, when market sentiment is excessively biased toward buying or selling, traders can take opposite operations in the hope that market sentiment will return to stability. Earn profit at the same time.
Martingale Strategy: A trading strategy based on probability theory, which compensates for the losses of previous transactions by increasing the amount invested in the next transaction, in the hope of making profits within a long enough time.
Grid Trading: A trading strategy that sets up a grid of buy and sell orders at a fixed price level in the hope of making profits from price fluctuations.
Elliott Wave Theory: A trading theory based on the laws of market fluctuations. It believes that market prices will change periodically according to certain fluctuation laws, and trading decisions are made by identifying and analyzing this fluctuation law.
Financial theory categories of training:
Markowitz Portfolio Theory: Proposed by Harry Markowitz, it describes the trade-off between risk and return in an investment portfolio and emphasizes reducing risk through diversified investments.
Beta Coefficient: It is used to measure the sensitivity of an asset relative to the overall market fluctuations and is widely used in portfolio risk assessment and asset pricing.
Capital Asset Pricing Model (CAPM): Proposed by William Sharp, John Lintner and Jack Traynor, it describes the relationship between asset expected return and risk, and is a measure of asset pricing and risk. base model.
Efficient Market Hypothesis (EMH): Proposed by Eugene Fama, it is believed that market prices already contain all available information, and investors cannot obtain excess profits through analysis.
Option Pricing Theory: includes models such as the Black-Skoll theorem, which is used to calculate the fair value of options and is of great significance to option pricing and risk management.
Linear Risk Model: A model used to quantify portfolio risk. It decomposes risk into a linear combination of different factors and helps with risk management and asset allocation.
Discrete-Time Option Pricing Models: including binary tree models, Monte Carlo simulations, etc., used to calculate option prices and arbitrage opportunities in discrete time.
Linear Option Pricing Models: such as the Black-Scholes option pricing model, which is used to calculate the price of European options and is the basis for pricing financial derivatives.
American Option Pricing Models: Used to calculate the price of options that can be executed at any time, taking into account the impact of early execution on option prices.
Implied Volatility Models: Used to estimate the implied volatility in option prices and quantify expected market volatility.
Stochastic Volatility Models: Models that describe stochastic changes in volatility, which have an important impact on financial derivatives pricing and risk management.
Mean Reversion Models: Models that believe asset prices will return to their equilibrium levels, used for short-term trading strategies and risk management.
Asset Pricing Multi-Factor Models: Models that relate asset returns to multiple influencing factors, such as the Fama-French three-factor model.
Financial Engineering: A subject area that applies techniques such as mathematics, statistics, and computer science to design financial products and trading strategies.
Technical Analysis: An analysis method that predicts future price trends through information such as historical prices and trading volumes. It is an investment decision-making method based on market behavior.
Fundamental Analysis: An analytical method that evaluates investment value by studying fundamental factors such as company financial statements and industry development trends.
Behavioral Finance: Study the impact of human behavior on financial market decisions, revealing the impact of factors such as investor sentiment and cognitive biases on the market.
Event Study: A research method that evaluates event value and market reaction by analyzing the impact of specific events on stock prices.
Arbitrage Pricing Theory (APT): describes the relationship between asset prices and a series of factors, and evaluates the value of assets by identifying arbitrage opportunities.
Arbitrage Theory: The theory of obtaining profits from market price differences by buying at low prices and selling at high prices. It is one of the common trading strategies in the financial market.
Related mathematical model classes for training:
Black-Scholes Model: An option pricing model used in the financial field, which describes the relationship between option prices and changes in underlying asset prices over time.
Markov Chain: A mathematical model that describes state transition in a random process, in which the future state only depends on the current state and has nothing to do with the past state.
Monte Carlo Method: A numerical calculation method based on random sampling, used to solve complex problems such as integration, optimization and probability simulation.
Percolation Theory: A mathematical theory that studies the permeability properties of liquid or gas in a medium, including permeability thresholds and percolation paths.
Brownian Motion: A mathematical model that describes the random motion of particles in a liquid or gas. It is widely used in fields such as finance, physics, and biology.
Complex Network Theory: A mathematical theory that studies the interconnection relationships between nodes and edges in complex systems, and is used to analyze social networks, neural networks, etc.
Graph Theory: A branch of mathematics that studies the structure and properties of graphs and is used to solve problems such as networks, path planning, and combinatorial optimization.
Linear Programming: A mathematical method for solving optimization problems through linear models, which is used in production planning, resource allocation and other fields.
Nonlinear Dynamics: The mathematical theory that studies the behavior of nonlinear systems, including concepts such as chaos and attractors.
Discrete Mathematics: A branch of mathematics that studies discrete objects and their properties, including set theory, graph theory, logic, etc.
Differential Equations: A branch of mathematics that studies the relationship between functions and their derivatives, and is used to describe the changes in natural phenomena.
Dynamic Programming: A mathematical method for solving multi-stage decision-making problems that improves efficiency by decomposing the problem into sub-problems and storing intermediate results.
Linear Algebra: A branch of mathematics that studies vector spaces and their transformations. It is widely used in fields such as machine learning and image processing.
Probability Theory: A branch of mathematics that studies random events and their probability distributions and is used to describe uncertainty and risk.
Functional Analysis: A branch of mathematics that studies function spaces and their properties, including Banach spaces, Hilbert spaces, etc.
Differential Geometry: Study of the differential properties of geometric objects such as curves and surfaces and their generalization to manifolds.
Discrete Optimization: A mathematical method for solving optimization problems of discrete variables, such as combinatorial optimization, integer programming, etc.
Wave Equation: A partial differential equation describing wave propagation behavior, used in fields such as acoustics and optics.
Matrix Theory: A branch of mathematics that studies matrices and their properties, including eigenvalues, eigenvectors, etc.
Control Theory: A mathematical theory that studies the control and stability of dynamic systems and is used to design controllers and optimize system performance.
Last updated