Qlib TabNet Configuration for CSI 300 Alpha158 Backtesting
Summary
This configuration defines a Qlib workflow for training a TabNet model on Alpha158 features for CSI 300 stocks, using Chinese market data and the CSI 300 index as benchmark. It sets a historical data window, separates fitting, validation, and test periods, applies robust feature normalization and missing-value filling, and rank-normalizes labels. The target is a short-horizon forward close-to-close return.
For portfolio analysis, the configuration uses a top-k dropout strategy that holds the highest-ranked signals and replaces a subset over time. It specifies a backtest window, account size, closing-price execution, transaction costs, and a price-limit threshold. Signal, long-short analysis, and portfolio analysis records are configured. This is a runnable setup description rather than a report of findings: it includes no model metrics, return results, or comparison against alternatives. Outcomes depend on the underlying data, library versions, and the assumptions encoded in the configuration.
Key ideas
- The workflow trains a TabNet model on Qlib’s Alpha158 features for CSI 300 stocks.
- Feature preprocessing includes robust normalization and filling missing values.
- The label represents a short-horizon forward return, while training and testing use separate date segments.
- Portfolio construction uses a top-k dropout strategy with transaction costs and a benchmark.
- The configuration provides no model-performance or backtest results.
Tags
Full text
# workflow_config_TabNet_Alpha158.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: TabnetModel
module_path: qlib.contrib.model.pytorch_tabnet
kwargs:
d_feat: 158
pretrain: True
seed: 993
dataset:
class: DatasetH
module_path: qlib.data.dataset
kwargs:
handler:
class: Alpha158
module_path: qlib.contrib.data.handler
kwargs: *data_handler_config
segments:
pretrain: [2008-01-01, 2014-12-31]
pretrain_validation: [2015-01-01, 2016-12-31]
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.