Qlib LightGBM Alpha360 Model and CSI 300 Backtest Setup
Summary
This configuration describes a Qlib workflow for training a LightGBM model on Alpha360 features and evaluating its predictions on CSI 300 constituents. The label compares closing prices across the next two reference periods. The data handler drops missing labels and rank-normalizes labels cross-sectionally; the dataset is divided into training, validation, and test periods, with fitting limited to the training span.
For portfolio evaluation, the configuration uses a top-k dropout strategy that holds up to 50 names and replaces up to five, with the CSI 300 index as benchmark. The backtest specifies closing-price execution, transaction costs, a minimum fee, and a price-limit threshold. Signal, signal-analysis, and portfolio-analysis records are enabled. These settings make the document useful as an example of an end-to-end equity modeling and backtesting workflow. It supplies configuration choices but no reported performance, robustness checks, or evidence that the model generalizes beyond the selected China-market data and dates.
Key ideas
- The workflow trains a LightGBM model using Alpha360 features for CSI 300 equities.
- The target is based on the ratio of closing prices across two future reference periods.
- Training, validation, and test data occupy distinct date ranges, while model fitting uses the training period.
- Portfolio evaluation uses a top-k dropout strategy and benchmarks results against the CSI 300 index.
- The backtest specifies closing-price execution, transaction costs, and a price-limit threshold.
Tags
Full text
# workflow_config_lightgbm_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: LGBModel
module_path: qlib.contrib.model.gbdt
kwargs:
loss: mse
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
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.