價格：免費
更新日期：2019-06-20
檔案大小：因裝置而異
目前版本：1.1.0
版本需求：Android 4.1 以上版本
官方網站：https://www.fundamentalspeculation.io
Email：support@fundamentalspeculation.io
聯絡地址：33 Glenwood Ave Apt 1, Norwalk CT 06854
Get all the latest Financial News, Social Media Chatter, access to some of the most Sophisticated Financial Models in the industry and the ability to build your own custom models all in one app.
Fundamental Speculation gives you access to our unique Relative Value Model which determines fair value based on a cohort of comparable companies with business fundamentals most similar to the target company in a large N-dimensional feature space of fundamental features. For more information, please see: https://fundamentalspeculation.io/model_rvm
We have trained our Deep Neural Network (DNN) to identify bullish/bearish patterns based on 20 years of price action and volume data from 1996-2016 over hundreds of stocks. For more information, please see: https://fundamentalspeculation.io/model_momentum
You will also be able to build your own sophisticated Discounted Cashflow Models, Relative Value Models, Scenario Analysis Models and Deep Learning Pattern Match Models.
Discounted Cashflow Models (DCF): We support single/multi-stage models with custom terminal values based on exit multiples (Price/Earnings, Price/Sales, Price/FCF). Our App will walk you though every step of the process allowing you to build a sophisticated model in simple, easy steps.
Relative Value Models: We allow you to build relative value models either by manually specifying the comparable companies or using our sophisticated algorithm to determine the cohort of companies. Our algorithm identifies the companies most similar to your target company in a large N-dimensional feature space of fundamental features specified by you. You will also pick the metrics you want to use to determine fair value.
Scenario Analysis Models: Scenario Analysis allows you to specify multiple Bull/Bear scenarios for the business fundamentals of the target company and our sophisticated algorithm will determine fair value metrics for each scenario and discount it back to a fair value for the company today based on your required rate of return.
Deep Learning Pattern Match Models: Neural Networks are a very powerful tool to find patterns in data. Build your own custom Deep Learning models to find patterns in price action and volume data of individual stocks in simple, easy to understand steps.
Supported Activation Functions: Linear, Sigmoid, Hard Sigmoid, TanH, ReLU, Softsign, Softplus, ELU, Softmax
Supported Optimizers: Stochastic gradient descent (SGD), RMSProp, Adagrad, Adadelta, Adam, Adamax, Nadam
Supported Loss Functions: Mean Squared Error, Mean Absolute Error, Mean Squared Logarithmic Error (MSLE), Log-Cosh, Poisson, Cosine Proximity, Binary Cross-Entropy