Qlib GRU Alpha360 Configuration for CSI 300 Stock Ranking
Summary
This configuration sets up a Qlib experiment that uses a gated recurrent unit model with Alpha360 features to rank CSI 300 constituents. Feature values are robustly normalized with outlier clipping and missing-value filling; labels use cross-sectional rank normalization after rows with missing labels are dropped. The target is a forward close-to-close return over the specified short horizon.
The configuration separates training, validation, and test periods, then evaluates a top-50 portfolio with periodic replacement of five holdings. Its backtest specifies the CSI 300 benchmark, close-price execution, a limit threshold, and transaction costs. These settings describe an experimental pipeline, not its outcome: the document contains no metrics, comparison against alternatives, or evidence that the model generalizes. Results would also depend on the data, execution assumptions, and implementation details.
Key ideas
- The experiment applies a GRU to Alpha360 features for CSI 300 instruments.\nRobust feature normalization and missing-value filling are configured before inference.\nThe label represents a forward close-to-close return, with cross-sectional rank normalization during learning.\nA top-50 strategy replaces five holdings at a time and is evaluated against the CSI 300 benchmark.\nThe configuration gives no performance results, so predictive value and robustness remain unknown.
Tags
Full text
# workflow_config_gru_Alpha360.yaml
```yaml
qlib_init:
provider_uri: "~/.qlib/qlib_data/cn_data"
region: cn
market: &market csi300
benchmark: &benchmark SH000300
data_handler_config: &data_handler_config
start_time: 2008-01-01
end_time: 2020-08-01
fit_start_time: 2008-01-01
fit_end_time: 2014-12-31
instruments: *market
infer_processors:
- class: RobustZScoreNorm
kwargs:
fields_group: feature
clip_outlier: true
- class: Fillna
kwargs:
fields_group: feature
learn_processors:
- class: DropnaLabel
- class: CSRankNorm
kwargs:
fields_group: label
label: ["Ref($close, -2) / Ref($close, -1) - 1"]
port_analysis_config: &port_analysis_config
strategy:
class: TopkDropoutStrategy
module_path: qlib.contrib.strategy
kwargs:
signal: <PRED>
topk: 50
n_drop: 5
backtest:
start_time: 2017-01-01
end_time: 2020-08-01
account: 100000000
benchmark: *benchmark
exchange_kwargs:
limit_threshold: 0.095
deal_price: close
open_cost: 0.0005
close_cost: 0.0015
min_cost: 5
task:
model:
class: GRU
module_path: qlib.contrib.model.pytorch_gru
kwargs:
d_feat: 6
hidden_size: 64
num_layers: 2
dropout: 0.0
n_epochs: 200
lr: 1e-3
early_stop: 20
batch_size: 800
metric: loss
loss: mse
GPU: 0
dataset:
class: DatasetH
module_path: qlib.data.dataset
kwargs:
handler:
class: Alpha360
module_path: qlib.contrib.data.handler
kwargs: *data_handler_config
segments:
train: [2008-01-01, 2014-12-31]
valid: [2015-01-01, 2016-12-31]
test: [2017-01-01, 2020-08-01]
record:
- class: SignalRecord
module_path: qlib.workflow.record_temp
kwargs:
model: <MODEL>
dataset: <DATASET>
- class: SigAnaRecord
module_path: qlib.workflow.record_temp
kwargs:
ana_long_short: False
ann_scaler: 252
- class: PortAnaRecord
module_path: qlib.workflow.record_temp
kwargs:
config: *port_analysis_config
```Shown in full with attribution under the source's licence. Licence: MIT
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.