Finamatrix A.I. Aggregator Architecture

August 26, 2017

The following is a summary of the principles and foundations of the Finamatrix AI Aggregator architecture and technology frameworks. Codes and programs of our various APIs will be summarized and provided in GitHub at https://github.com/finamatrix. Core optimization functions that are at the heart of our tech will only be summarily provided here in the following research papers. Our technology is mature and many enhancements have been performed since 2004.

Atomic Portfolio Selection: MVSK Utility Optimization

Dr Lanz Chan, CEO Finamatrix
(Supervised by Professor KH Liow)

Date Written: June 16, 2004

This research utilizes higher-moment risk-return relationships and portfolio selection strategies. It determines whether higher moments are significantly priced and evaluates time-varying high-moment characteristics of assets. A higher-moment (mean, variance, skewness, kurtosis or MVSK) portfolio selection framework is suggested and performance results are compared with the standard mean-variance (MV) method.

Keywords: co-variance, co-skewness, co-kurtosis, anti-moment, bi-moments, tri-moments, time-varying Kalman Filter.

Reference:

Chan, Lanz, Atomic Portfolio Selection: MVSK Utility Optimization of Global Real Estate Securities (June 16, 2004). Finamatrix, July 2011. Available at SSRN: https://ssrn.com/abstract=1744802

Automated Trading with Genetic-Algorithm Neural-Network Risk-Cybernetics

Dr Lanz Chan, CEO Finamatrix
(Supervised by Professor WK Wong)

Date Written: February 20, 2012

Recent years have witnessed the advancement of automated algorithmic trading systems as institutional solutions in the form of autobots, black box or expert advisors. However, little research has been done in this area with sufficient evidence to show the efficiency of these systems. This paper builds an automated trading system which implements an optimized genetic-algorithm neural-network (GANN) model with cybernetic concepts and evaluates the success using a modified value-at-risk (MVaR) framework. The cybernetic engine includes a circular causal feedback control feature and a developed golden-ratio estimator, which can be applied to any form of market data in the development of risk-pricing models. It is shown that the technique is useful as a trading and volatility control system for institutions including central bank monetary policy as a risk-minimizing strategy. Furthermore, the results are achieved within a 30-second timeframe for an intra-week trading strategy, offering relatively low latency performance. The results show that risk exposures are reduced by four to five times with a maximum possible success rate of 96%, providing evidence for further research and development in this area.

Keywords: Automation, Autobot, Genetic-Algorithm Neural-Network, Risk-Pricing, Risk Cybernetics, Expert Advisor.

Reference:

Chan, Lanz and Wong, Wing-Keung, Automated Trading with Genetic-Algorithm Neural-Network Risk Cybernetics: An Application on FX Markets (February 20, 2012). Finamatrix Journal, February 2012 . Available at SSRN: https://ssrn.com/abstract=1687763

Robust Prediction in Nearly Periodic Time Series Using Motifs

Neural Networks (IJCNN), 2014 International Joint Conference
Woon Huei Chai, PhD Candidate, CTO Finamatrix

Date Written: July 06, 2014

We consider the prediction task for a process with nearly periodic property, i.e. patterns occur with some regularities but no exact periodicity. We propose an inference approach based on probabilistic Markov framework utilizing motif-driven transition probabilities for sequential prediction. In particular, a Markov-based weighting framework utilizing fully the information from recent historical data and sequential pattern regularities is developed for nearly periodic time series prediction. Preliminary experimental results show that our prediction approach is competitive against the moving average and multi-layer perceptron neural network approaches on synthetic data. Moreover, our proposed method is shown to be empirically robust on time-series with missing data and noise. We also demonstrate the usefulness of our proposed approach on a real-world vehicle parking lot availability prediction task.

Keywords: Time series analysis, Predictive models, Markov processes, Prediction algorithms, Robustness, Neural networks, Noise.

Reference:

Chai, Woon Huei, Robust Prediction in Nearly Periodic Time Series Using Motifs (July 06, 2014). Neural Networks (IJCNN), 2014 International Joint Conference . Available at SSRN: https://ssrn.com/abstract=3007625

A Fast Sparse Reconstruction Approach for High Resolution Image-Based Object Surface Anomaly Detection

Machine Vision Applications (MVA), 2017 Fifteenth IAPR International Conference
Woon Huei Chai, PhD Candidate, CTO Finamatrix

Date Written: May 8, 2017

We propose an approach to resolve two issues in a recent proposed sparse reconstruction based, anomaly detection approach as a part of automated visual inspection (AVI). The original approach needs large computation and memory for high resolution problem.

To solve it, we proposed a two-step sparse reconstruction:

1) the first sparse representation of input image is estimated in a sparse reconstruction with low resolution down-sampled images and,

2) the high resolution residual values is generated in another sparse reconstruction with the sparse representation.

The first step provides the flexibility of freely adjusting the computation and the demand of memory storage with small trade-off of detection accuracy. Moreover, an illumination adaptive threshold with morphological operators is used in the anomaly classification. Empirical results show that the proposed approach can effectively replace the original approach with better results.

Reference:

Chai, Woon Huei, A Fast Sparse Reconstruction Approach for High Resolution Image-Based Object Surface Anomaly Detection (May 8, 2017). Machine Vision Applications (MVA), 2017 Fifteenth IAPR International Conference, ISBN: 978–4–9011–2216–0 . Available at SSRN: https://ssrn.com/abstract=3007646