Qlib ADARNN Configuration for CSI 300 Stock Prediction
Summary
This configuration specifies a Qlib workflow for training an ADARNN model on CSI 300 equities in the Chinese market. It uses Alpha360 features, robust z-score normalization with outlier clipping, feature filling for missing values, and cross-sectional rank normalization for labels. The target is a forward close-price return defined using the next two reference points. Data is divided into training, validation, and test periods, followed by signal analysis and portfolio evaluation.
For portfolio analysis, the setup uses a top-k dropout strategy that holds 50 names and drops five, with the CSI 300 index as benchmark. The backtest specifies close-price dealing, transaction costs, a minimum fee, and a limit threshold. The neural model uses six input features, two layers, a hidden size of 64, mean squared error loss, and early stopping. This is an experiment recipe rather than a report: it includes no prediction or portfolio results, and its outcomes depend on the data, implementation, and execution assumptions. Its specified historical intervals also do not establish performance beyond the test period.
Key ideas
- The workflow trains ADARNN on Alpha360 features for CSI 300 instruments.
- Features are robustly normalized and missing values are filled, while labels receive cross-sectional rank normalization.
- The target represents a forward close-price return, and the data is split into training, validation, and test segments.
- Portfolio evaluation uses a top-k dropout strategy with a CSI 300 benchmark and explicit transaction-cost assumptions.
- The configuration describes an experiment but reports no model accuracy or investment performance.
Tags
Full text
# workflow_config_adarnn_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: ADARNN
module_path: qlib.contrib.model.pytorch_adarnn
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.