Bitcoin Time Series Forecasting Model Results

Last updated on: 2019-03-19

Alan's Page

In no event shall the author be liable for any direct, indirect, incidental, special, consequential or exemplary damages including damages for loss of revenue, anticipated profits, loss of business, goodwill, or other tangible or intangible losses resulting from the use of any data or information provided in this article.

The 1st most accurate forecasting model is the Weighted Average model.

Regression model weight = 0.02 ; ARIMA model weight = 0.98

The forcasted bitcoin prices for the coming 5 days are:
Lower.99..CI Lower.95..CI Lower.68..CI Forecasted.value Upper.68..CI Upper.95..CI Upper.99..CI
11 3261 3447 3738 4038 4338 4630 4816
12 2953 3212 3618 4036 4454 4861 5120
13 2744 3053 3538 4038 4538 5023 5332
14 2553 2908 3464 4037 4610 5166 5521
15 2394 2787 3403 4038 4672 5288 5681
model ME MAE MSE MPE MAPE
3 Weighted Average 0.66 44 2590 0 1.1

The 2nd most accurate forecasting model is the AutoARIMA model.

ARIMA(p,d,q,P,D,Q)[m]: p is 1 , q is 1 , P is 0 , Q is 0 , m is 1 , d is 1 , D is 0; Coefficient ar1 is -0.7 ; Coefficient ma1 is 0.77

The forcasted bitcoin prices for the coming 5 days are:
Lower.99..CI Lower.95..CI Lower.68..CI Forecasted.value Upper.68..CI Upper.95..CI Upper.99..CI
6 3305 3465 3716 3974 4232 4483 4643
7 2991 3225 3593 3972 4350 4718 4952
8 2778 3064 3512 3973 4435 4883 5168
9 2583 2915 3436 3972 4508 5029 5361
10 2422 2793 3374 3973 4571 5153 5524
model ME MAE MSE MPE MAPE
2 AutoARIMA 69 69 7339 1.7 1.7

The 3rd most accurate forecasting model is the Regression model.

Fitted regression equation is: price^ = -1549.04 + 5.83 *t R-squared = 0.49 p-value = 0

The forcasted bitcoin prices for the coming 5 days are:
Lower.99..CI Lower.95..CI Lower.68..CI Forecasted.value Upper.68..CI Upper.95..CI Upper.99..CI
835 2457 4995 7607 10219 12757 14379
841 2463 5001 7613 10225 12763 14384
847 2468 5007 7619 10231 12769 14390
853 2474 5013 7624 10236 12775 14396
859 2480 5018 7630 10242 12781 14402
model ME MAE MSE MPE MAPE
Regression -3766 3766 1.4e+07 -96 96