Mlp

Here, we present the model's performance in predicting the bitcoin price along with their hyper-parameters. For more details on the model, see here.

Prediction

Prices

DATE	CLOSE
2024-12-18	98,220.27
2024-12-25	100,050.01
2025-01-01	101,157.45
2025-01-08	101,368.88

Charts

Model Performance

For scoring, Root Mean Squared Error was used.

Training data

Mean across all targets: 2,390.32

D+7	D+14	D+21	D+28
2,083.20	2,122.15	1,537.67	3,818.27

Test data

Mean across all targets: 5,685.46

D+7	D+14	D+21	D+28
4,789.17	6,129.12	7,632.03	4,191.50

Model Setup

Below are choices made for the model setup.

Exogenous Data

(Determined by Forward Selection and heuristically based on the model's assumptions)

param	description	value
include_indices	Whether to include market indices.	False
include_commodities	Whether to include commodities.	False
include_currencies	Whether to include currencies.	False

Feature Generator

(Determined by Grid Search)

param	description	value
ohlc_ratio	Whether to add OHLC ratios of bitcoin price to the data.	True
lags	List of lags to use. None, if no lags are used.	None

Feature Scaler

(Determined by heuristically, considering the model's assumptions)

param	description	value
log	Whether to take the log of the data.	True
diff	Whether to difference the data.	False
normalize	Whether to normalize the data.	True

Hyper-parameters

(Determined by Grid Search)

param	description	value
fc_dims	List of fully connected layer dimensions.	[20, 15, 10]
dropout_rates	List of dropout rates.	[0.2, 0.2, 0.2]