Regression

ChickWeights

Summary

Model	MAE	RMSE	R2	Memory in Mb	Time in s
Deep River Attention	79.3499	93.1322	-0.720105	0.0645237	134.626
Deep River LSTM	100.133	117.539	-1.7398	0.0351133	138.265
Deep River Linear	119.58	138.47	-2.80249	0.0177469	20.0831
Deep River MLP	24.8823	38.4367	0.707013	0.0374966	29.0197
Deep River RNN	100.316	117.739	-1.74914	0.0329514	54.5518
Linear regression	24.488	37.1301	0.726594	0.00423336	2.12884
River MLP	121.818	141.004	-2.94294	0.0126944	21.2278
[baseline] Mean predictor	50.2509	71.1144	-0.00292947	0.000490189	0.755343

Charts

TrumpApproval

Summary

Model	MAE	RMSE	R2	Memory in Mb	Time in s
Deep River Attention	6.49609	13.1981	-58.5337	0.067049	189.508
Deep River LSTM	10.7062	16.3447	-90.3047	0.0376387	194.99
Deep River Linear	36.6534	36.7298	-460.078	0.0187769	31.5477
Deep River MLP	1.31522	4.85733	-7.06366	0.0385265	45.7201
Deep River RNN	10.936	16.4744	-91.7587	0.0354767	88.6672
Linear regression	1.62028	4.53607	-6.0323	0.0049963	3.96432
River MLP	5.50408	10.9984	-40.3426	0.0139608	37.746
[baseline] Mean predictor	1.56755	2.20286	-0.658483	0.000490189	1.44011

Charts

Datasets

ChickWeights

Chick weights along time.

The stream contains 578 items and 3 features. The goal is to predict the weight of each chick along time, according to the diet the chick is on. The data is ordered by time and then by chick.

Name  ChickWeights                                                                                                          
Task  Regression

Samples 578
Features 3
Sparse False
Path /Users/cedrickulbach/Documents/Projects/deep-river/.venv/lib/python3.10/site-packages/river/datasets/chick-weights.csv

TrumpApproval

Donald Trump approval ratings.

This dataset was obtained by reshaping the data used by FiveThirtyEight for analyzing Donald Trump's approval ratings. It contains 5 features, which are approval ratings collected by 5 polling agencies. The target is the approval rating from FiveThirtyEight's model. The goal of this task is to see if we can reproduce FiveThirtyEight's model.

Name  TrumpApproval                                                                                                             
Task  Regression

Samples 1,001
Features 6
Sparse False
Path /Users/cedrickulbach/Documents/Projects/deep-river/.venv/lib/python3.10/site-packages/river/datasets/trump_approval.csv.gz

Models

Linear regression

Pipeline (
  StandardScaler (
    with_std=True
  ),
  LinearRegression (
    optimizer=SGD (
      lr=Constant (
        learning_rate=0.005
      )
    )
    loss=Squared ()
    l2=0.
    l1=0.
    intercept_init=0.
    intercept_lr=Constant (
      learning_rate=0.01
    )
    clip_gradient=1e+12
    initializer=Zeros ()
  )
)

Deep River Linear

Pipeline (
  StandardScaler (
    with_std=True
  ),
  LinearRegressionInitialized (
    n_features=10
    loss_fn="mse"
    optimizer_fn="sgd"
    lr=0.005
    is_feature_incremental=True
    device="cpu"
    seed=42
    gradient_clip_value=None
  )
)

Deep River MLP

Pipeline (
  StandardScaler (
    with_std=True
  ),
  MultiLayerPerceptron (
    n_features=10
    n_width=5
    n_layers=5
    loss_fn="mse"
    optimizer_fn="sgd"
    lr=0.005
    is_feature_incremental=True
    device="cpu"
    seed=42
    gradient_clip_value=None
  )
)

Deep River LSTM

Pipeline (
  StandardScaler (
    with_std=True
  ),
  LSTMRegressor (
    n_features=10
    hidden_size=64
    num_layers=1
    dropout=0.1
    gradient_clip_value=1.
    loss_fn="mse"
    optimizer_fn="adam"
    lr=0.001
    is_feature_incremental=True
    device="cpu"
    seed=42
  )
)

Deep River RNN

Pipeline (
  StandardScaler (
    with_std=True
  ),
  RNNRegressor (
    n_features=10
    hidden_size=64
    num_layers=1
    nonlinearity="tanh"
    dropout=0.1
    gradient_clip_value=1.
    loss_fn="mse"
    optimizer_fn="adam"
    lr=0.001
    is_feature_incremental=True
    device="cpu"
    seed=42
  )
)

River MLP

Pipeline (
  StandardScaler (
    with_std=True
  ),
  MLPRegressor (
    hidden_dims=(10,)
    activations=(, , )
    loss=Squared ()
    optimizer=SGD (
      lr=Constant (
        learning_rate=0.005
      )
    )
    seed=42
  )
)

[baseline] Mean predictor

StatisticRegressor (
  statistic=Mean ()
)

Environment

Python implementation: CPython
Python version       : 3.12.12
IPython version      : 9.6.0

river       : 0.22.0
numpy       : 1.26.4
scikit-learn: 1.5.2
pandas      : 2.2.3
scipy       : 1.16.2

Compiler    : Clang 21.1.4 
OS          : Linux
Release     : 6.11.0-1018-azure
Machine     : x86_64
Processor   : x86_64
CPU cores   : 4
Architecture: 64bit