Qlib Double-Ensemble Workflow for CSI 300 Stock Ranking
Summary
This Qlib configuration describes a machine-learning workflow for ranking CSI 300 stocks. It uses the Alpha360 data handler and a double-ensemble model built from gradient-boosted trees. The configured label is a forward close-price return, while the learning pipeline drops missing labels and cross-sectionally rank-normalizes them. Data is divided into training, validation, and test periods, with the portfolio backtest beginning in the test period.
For portfolio construction, the workflow holds the top 50 signals and allows five holdings to be dropped as rankings change. It specifies the CSI 300 index as benchmark, closing-price execution, transaction costs, and a price-limit threshold. Signal analysis and portfolio analysis records are configured, but the document contains no resulting metrics or conclusions. This is an experiment specification rather than evidence of profitability; results would depend on data quality, assumptions about execution and costs, and the model’s out-of-sample behavior.
Key ideas
- The workflow uses Qlib’s Alpha360 handler and a double-ensemble gradient-boosted model.
- Its target is a forward close-price return, with labels rank-normalized across stocks.
- The dataset is divided into training, validation, and test segments.
- A top-ranked portfolio uses a dropout rule to adjust holdings as signals change.
- The configuration specifies benchmark, execution, and transaction-cost assumptions but reports no performance results.
Tags
Full text
# workflow_config_doubleensemble_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: []
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: DEnsembleModel
module_path: qlib.contrib.model.double_ensemble
kwargs:
base_model: "gbm"
loss: mse
num_models: 3
enable_sr: True
enable_fs: True
alpha1: 1
alpha2: 1
bins_sr: 10
bins_fs: 5
decay: 0.5
sample_ratios:
- 0.8
- 0.7
- 0.6
- 0.5
- 0.4
sub_weights:
- 1
- 1
- 1
epochs: 136
colsample_bytree: 0.8879
learning_rate: 0.0421
subsample: 0.8789
lambda_l1: 205.6999
lambda_l2: 580.9768
max_depth: 8
num_leaves: 210
num_threads: 20
verbosity: -1
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.