Alyssa Edge Adaptive — Self-Learning (Local)

Trade journal (R-based)

This stores on your phone/browser only (LocalStorage). Every time you add trades, the Adaptive engine updates automatically.

Date/Time

Asset

Side

Setup

Regime

Volatility state

Outcome (R)

Positive = win. Negative = loss. 0 = scratch.

Notes (optional)

Decision support only No wallet No execution No guarantees

“Self-learning” here means: it updates probabilities and weightings from your logged trades. It does not predict markets and it does not place trades.

Performance summary

Trades

0

count

Win rate

0%

R > 0

Avg R

0.00

mean

Expectancy

0.00

E[R]

Date	Asset	Side	Setup	Regime	R

Tip: Use Import/Export to move this data between devices.

Adaptive (Self-learning)

Two layers: (1) Bayesian per-setup win probability (Beta posterior) and (2) Online logistic model (SGD) for “win vs loss” probability using your features. Everything updates instantly when you add trades.

Bayesian: Setup / Regime

Best setup

—

by posterior win%

Confidence

—

sample size aware

Setup	Regime	Trades	Posterior win%	Avg R

Model: Live probability

Try a hypothetical trade and see what your learned model thinks.

Asset Side Setup Regime Volatility

Model win probability

—

learned from your data

Signal quality

—

based on sample size

This is not “AI that prints money.” It’s a disciplined feedback loop: your rules + your logs → better weighting.

Import / Export

Export a single JSON file of trades + model state. Import restores it on this device.

Raw JSON (view)

If you want this to be “public,” host this static site on Cloudflare Pages and share the link. If you want on-chain proof later, you can publish periodic “checkpoint hashes” from the head-hash shown in Registry.

What “self-learning” means here

It learns your edge from your own trade outcomes (R).
Bayesian layer makes win-rate estimates that don’t get fooled by low sample sizes.
Online model updates weights after every trade (like a thermostat, not like ChatGPT).
Local-only: nothing is sent to a server. You control the data.

No financial advice

This tool is for journaling and analytics. It does not execute trades and does not guarantee any performance.