Learnware & Reuser#

Learnware is the most basic concept in the learnware paradigm. In this section, we will introduce the concept and design of Learnware and its extension for Hetero Reuse. Then we will introduce the Reuse Methods, which applies one or several Learnwares to solve the user’s task.

Concepts#

In the learnware paradigm, a learnware is a well-performed trained machine learning model with a specification which enables it to be adequately identified to reuse according to the requirement of future users who know nothing about the learnware in advance. The introduction of specifications are shown in COMPONENTS: Specification.

In our implementation, the class Learnware has 3 important member variables:

  • id: The learnware id is generated by market.

  • model: The model in the learnware, can be a BaseModel or a dict including model name and path. When it is a dict, the function Learnware.instantiate_model is used to transform it to a BaseModel. The function Learnware.predict use the model to predict for an input X. See more in COMPONENTS: Model.

  • specification: The specification including the semantic specification and the statistic specification.

Learnware for Hetero Reuse#

In the Hetero Market(see COMPONENTS: Hetero Market for details), HeteroSearcher identifies and recommends helpful learnwares among all learnwares in the market, including learnwares with feature/label spaces different from the user’s task requirements(heterogeneous learnwares). FeatureAlignLearnware and HeteroMapLearnware are designed to enable the reuse of heterogeneous learnwares, which extends Learnware with the ability to align the feature space and label space of the learnware to the user’s task requirements, and provide basic interfaces for heterogeneous learnwares to be applied to tasks beyond their original purposes.

FeatureAlignLearnware#

FeatureAlignLearnware employs a neural network to align the feature space of the learnware to the user’s task. It is initialized with a Learnware, and has the following methods to expand the applicable scope of this Learnware:

  • align: Trains a neural network to align user_rkme, which is the RKMETableSpecification of the user’s data, with the learnware’s statistical specification.

  • predict: Predict the output for user data using the trained neural network and the original learnware’s model.

HeteroMapAlignLearnware#

If user data is not only heterogeneous in feature space but also in label space, HeteroMapAlignLearnware uses the help of a small amount of labeled data (x_train, y_train) required from the user task to align heterogeneous learnwares with the user task. There are two key interfaces in HeteroMapAlignLearnware:

  • HeteroMapAlignLearnware.align(self, user_rkme: RKMETableSpecification, x_train: np.ndarray, y_train: np.ndarray)

    • input space alignment: Align the feature space of the learnware to the user task’s statistical specification user_rkme using FeatureAlignLearnware.

    • output space alignment: Further align the label space of the aligned learnware to the user task through supervised learning of FeatureAugmentReuser using (x_train, y_train).

  • HeteroMapAlignLearnware.predict(self, user_data)

    • If input space and output space alignment are both performed, use the FeatureAugmentReuser to predict the output for user_data.

All Reuse Methods#

In addition to applying Learnware, FeatureAlignLearnware or HeteroMapAlignLearnware objects directly by calling their predict interface, the learnware package also provides a set of Reuse Methods for users to further customize a single or multiple learnwares, with the hope of enabling learnwares to be helpful beyond their original purposes, and eliminating the need for users to build models from scratch.

There are two main categories of Reuse Methods: (1) direct reuse and (2) reuse based on a small amount of labeled data.

Note

Combine HeteroMapAlignLearnware with the following reuse methods to enable the reuse of heterogeneous learnwares. See WORKFLOW: Hetero Reuse for details.

Direct Reuse of Learnware#

Two methods for direct reuse of learnwares are provided: JobSelectorReuser and AveragingReuser.

JobSelectorReuser#

JobSelectorReuser trains a classifier job selector that identifies the optimal learnware for each data point in user data. There are three member variables:

  • learnware_list: A list of Learnware objects for the JobSelectorReuser to choose from.

  • herding_num: An optional integer that specifies the number of items to herd, which defaults to 1000 if not provided.

  • use_herding: A boolean flag indicating whether to use kernel herding.

The most important methods of JobSelectorReuser are job_selector and predict:

  • job_selector: Train a job selector based on user’s data and the learnware_list. Processions are different based on the value of use_herding:

    • If use_herding is False: Statistical specifications of learnwares in learnware_list combined with the corresponding learnware index are used to train the job selector.

    • If use_herding is True:

      • Estimate the mixture weight based on user raw data and the statistical specifications of learnwares in learnware_list

      • Use the mixture weight to generate herding_num auxiliary data points which mimic the user task’s distribution through the kernel herding method

      • Finally learns the job selector on the auxiliary data points.

  • predict: The job selector is essentially a multi-class classifier \(g(\boldsymbol{x}):\mathcal{X}\rightarrow \mathcal{I}\) with \(\mathcal{I}=\{1,\ldots, C\}\), where \(C\) is the size of learnware_list. Given a testing sample \(\boldsymbol{x}\), the JobSelectorReuser predicts it by using the \(g(\boldsymbol{x})\)-th learnware in learnware_list.

AveragingReuser#

AveragingReuser uses an ensemble method to make predictions. It is initialized with a list of Learnware objects, and has a member variable mode which specifies the ensemble method(default is set to mean).

  • predict: The member variable mode provides different options for classification and regression tasks:

    • For regression tasks, mode should be set to mean. The prediction is the average of the learnwares’ outputs.

    • For classification tasks, mode has two available options. If mode is set to vote_by_label, the prediction is the majority vote label based on learnwares’ output labels. If mode is set to vote_by_prob, the prediction is the mean vector of all learnwares’ output label probabilities.

Reuse Learnware with Labeled Data#

When users have a small amount of labeled data available, the learnware package provides two methods: EnsemblePruningReuser and FeatureAugmentReuser to help reuse learnwares. They are both initialized with a list of Learnware objects learnware_list, and have different implementations of fit and predict methods.

EnsemblePruningReuser#

The EnsemblePruningReuser class implements a selective ensemble approach inspired by the MDEP algorithm, as detailed in 1. It selects a subset of learnwares from learnware_list, utilizing user’s labeled data for effective ensemble integration on user tasks. This method effectively balances validation error, margin ratio, and ensemble size, leading to a robust and optimized selection of learnwares for task-specific ensemble creation.

  • fit: Effectively prunes the large set of learnwares learnware_list by evaluating and comparing the learnwares based on their performance on user’s labeled validation data (val_X, val_y). Returns the most suitable subset of learnwares.

  • predict: The mode member variable has two available options. Set mode to regression for regression tasks, and classification for classification tasks. The prediction is the average of the selected learnwares’ outputs.

FeatureAugmentReuser#

FeatureAugmentReuser helps users reuse learnwares by augmenting features. In this method, outputs of the learnwares from learnware_list on user’s validation data val_X are taken as augmented features and are concatenated with original features val_X. The augmented data(concatenated features combined with validation labels val_y) are then used to train a simple model augment_reuser which gives the final prediction on user_data.

  • fit: Trains the augment_reuser using augmented user validation data. For classification tasks, mode should be set to classification, and augment_reuser is a LogisticRegression model. For regression tasks, mode should be set to classification, and augment_reuser is a RidgeCV model.

References#

1

Yu-Chang Wu, Yi-Xiao He, Chao Qian, and Zhi-Hua Zhou. Multi-objective Evolutionary Ensemble Pruning Guided by Margin Distribution. In: Proceedings of the 17th International Conference on Parallel Problem Solving from Nature (PPSN’22), Dortmund, Germany, 2022.