I can contribute 20 hours a week to work on a trading AI for the next 10 years
Nov 28, 2022 at 07:40
402 개의 뷰
1 Replies
Nov 28, 2022 at 07:40
Hi
I'm an Australian disability pensioner and I will start ten years of (part time) study for a degree in Mathematics and AI next February. I want to start; or join a team, that will work on; or is already working on, producing a suite of cloud applications that train and evolve AI's to trade one major against all the other majors (seeing that I am Australian I would favor working to produce an AUD expert system). At the moment I know very little about AI and even less about the currency markets. I'm playing around with genetic algorithms that evolve neural networks to play video games and I have yet to come across any literature regarding deep learning (which I have been told by a PhD candidate that I know is the shizzle). I am confident that I can evolve a group of break even networks that trade the AUD based on technical indicators in the coming year or two and in the long term I want to create a group of networks that can decode news events into a collection of technical indicators that could then be fed into new fuzzy logic networks along with a range of truly technical indicators.
As I said in the topic I can contribute 20 hours a week for free. Can anyone recommend a GitHub project that even remotely resembles what I am talking about?
I'm an Australian disability pensioner and I will start ten years of (part time) study for a degree in Mathematics and AI next February. I want to start; or join a team, that will work on; or is already working on, producing a suite of cloud applications that train and evolve AI's to trade one major against all the other majors (seeing that I am Australian I would favor working to produce an AUD expert system). At the moment I know very little about AI and even less about the currency markets. I'm playing around with genetic algorithms that evolve neural networks to play video games and I have yet to come across any literature regarding deep learning (which I have been told by a PhD candidate that I know is the shizzle). I am confident that I can evolve a group of break even networks that trade the AUD based on technical indicators in the coming year or two and in the long term I want to create a group of networks that can decode news events into a collection of technical indicators that could then be fed into new fuzzy logic networks along with a range of truly technical indicators.
As I said in the topic I can contribute 20 hours a week for free. Can anyone recommend a GitHub project that even remotely resembles what I am talking about?
Nov 28, 2022 at 21:34
I have been thinking about what I want to do and seeings I estimate that there is something like a 60% chance that several major currencies are going to go into inflationary spirals before the end of the decade I would want to be set up to trade them against gold. After that I was thinking about testing the varsity of the various nations CPI's by plotting their reported inflation across currencies and seeing if they were divergent when grouped in sets of the same currency. I also wondered how inflation/deflation would chart from country to country when measured against the gold equivalent of the currency required to buy a standard basket of goods and what that would actually mean?
Here is the stuff of six monitors...
XAU/USD
XAU/EUR
XAU/GBP
XAU/JPY
XAU/CHF
XAU/CAD
To finish up I'm going to blow some hot air as I set of from base camp on my first attempt to climb Mount Stupid. I'm pondering that maybe I could evolve large sets of unique, individual, deep neural networks; each member of which would cast price predictions several time frames into the future. We could then map a line of best fit into that future and translate the the standard deviation boundaries of our own bell curve onto the chart.
Here is the stuff of six monitors...
XAU/USD
XAU/EUR
XAU/GBP
XAU/JPY
XAU/CHF
XAU/CAD
To finish up I'm going to blow some hot air as I set of from base camp on my first attempt to climb Mount Stupid. I'm pondering that maybe I could evolve large sets of unique, individual, deep neural networks; each member of which would cast price predictions several time frames into the future. We could then map a line of best fit into that future and translate the the standard deviation boundaries of our own bell curve onto the chart.
*상업적 사용 및 스팸은 허용되지 않으며 계정이 해지될 수 있습니다.
팁: 이미지/유튜브 URL을 게시하면 게시물에 자동으로 삽입됩니다!
팁: @기호를 입력하여 이 토론에 참여하는 사용자 이름을 자동으로 완성합니다.
