Nvidia earlier this month released cuDNN, a set of optimized low-level primitives to boost the processing speed of deep neural networks (DNN) on CUDA compatible GPUs. The company intends to help developers harness the power of graphics processing units for deep learning applications...
Simplex optimization is one of the simplest algorithms available to train a neural network. Understanding how simplex optimization works, and how it compares to the more commonly used back-propagation algorithm, can be a valuable addition to your machine learning skill set...
This paper exploits a novel hand collected dataset to provide a comprehensive analysis of the demographics and social relationships behind illegal insider trading networks...
“Information is the oil of the 21st century, and analytics is the combustion engine.” – Peter Sonergaard from Gartner Research...
WEEKLY DIGEST 2014, October 05 - 12 for Quantitative Analysis, Trading and Developing - mql5 digest Quantitative Finance: Analysis of Gold Mining ETF - 5 Reasons Why the Market Vectors Gold Mining ETF Could Be The Next Monster Trade...
Many of you who wished to use the Matlab’s Financial Toolbox for portfolio visualization most probably realised that something was wrong. Not every time the annualised risk and return for different time-series matched with what was expected...
With every year the competition in the field of quantitative finance gets tougher...
All the time in the displaying of uncommon occasions in money and protection industry the investigators are keen on both the likelihood of occasions and their monetary results. That rotates around the part of a likelihood of ruin, i.e...
FX trading styles: Different strokes for different folks The short term trader is also not immune from the competition that bank and hedge fund professionals offer, but has a greater margin of error...
Limit orders provide investors with an opportunity to buy or sell a specific number of shares at a given/demanded price. In many instances this solution is favored to market orders because of a slightly higher degree of control over risk or potential losses. We need to include it in estimations...
In my previous post I provided an introduction to the trading model invention and design. We made use of FX data of AUDUSD pair sampled hourly and splitting data into weekly time-series. Today, we will work a bit harder over formulation of the very first rules for the model...
Feeding our models with data for backtesting purposes requires pulling data from the server...
When you think about an invention of a new model for algorithmic trading, there are only three key elements you need to start your work with: creativity, data, and programming tool. Assuming that the last two are already in your possession, all what remains is seeking and finding a great new idea...
Dealing with financial data sets in a raw format sometimes we find ourselves in a need of converting a specific date and time into continuous measure of time expressed by a single number...
The tick-data provided in the .csv (comma separated values) file format sometimes may be a real problem to handle quickly, especially when the total size starts to count in hundreds of GB. If your goal is to extract a time-series with, say, hourly time resolution only, Read more -> Sources...
function [str]=bankformat(num) num2=floor(num*1000)/1000; r=int32(100*(num-floor(num))); str = num2str(num2); k=find(str == '.'); if(isempty(k)) str=[str,'.00']; end FIN = min(length(str),find(str == '...
A precious lesson I learned during my venture over programming an independent backtesting engine for new trading model was slippage. Simply speaking, slippage is a fraction of stock price which you need to assume as a deviation from the price you are willing to pay...
It is one of the most fundamental approaches in measuring the risk, but truly worth revising its calculation. Value-at-Risk (VaR...
When I read the book of Nassim Nicholas Taleb Black Swan my mind was captured by the beauty of extremely rare events and, concurrently, devastated by the message the book sent: the non-computability of the probability of the consequential rare events using scientific methods (as owed to the very...
Having in mind the upcoming series of articles on building a backtesting engine for algo traded portfolios, today I decided to drop a short post on a simulation of the portfolio realised profit and loss (P&L...
