Shulou(Shulou.com)06/01 Report--

How to carry out the functional analysis of Algobase () class, I believe that many inexperienced people do not know what to do about it. Therefore, this paper summarizes the causes and solutions of the problem. Through this article, I hope you can solve this problem.

The AlgoBase () class, as we mentioned earlier, is the parent of all algorithms, so AlgoBase () should abstract the common methods of all algorithms. In other words, the method in AlgoBase () is a function that all subclasses may have.

So let's take a look at what functions there are, and then pick some of the more important ones and take a look at the code.

Briefly introduce their functions.

Fit ()

Fit () We see in KNN's algorithm that the subclass also overrides this function, so what do you mainly do in the parent class? The main role in the parent class is to bind the corresponding trainset to the self, that is, to assign a self.trainset attribute to the current algorithm object.

Predict () def predict (self, uid, iid, r_ui=None, clip=True, verbose=False): # Convert raw ids to inner ids try: iuid = self.trainset.to_inner_uid (uid) except ValueError: iuid = 'UKN__' + str (uid) try: iiid = self.trainset.to_inner_iid (iid) except ValueError: iiid =' UKN__' + str (iid)

Details = {} try: est = self.estimate (iuid, iiid)

# If the details dict was also returned if isinstance (est, tuple): est, details = est

Details ['was_impossible'] = False

Except PredictionImpossible as e: est = self.default_prediction () details ['was_impossible'] = True details [' reason'] = str (e)

# clip estimate into [lower_bound, higher_bound] if clip: lower_bound, higher_bound = self.trainset.rating_scale est = min (higher_bound, est) est = max (lower_bound, est)

Pred = Prediction (uid, iid, r_ui, est, details)

If verbose: print (pred)

Return pred

We can simply look at the code into three parts. The first part is the exception handling of two try, which converts the raw id in the data set into the internal ID after processing. In the previous dataset processing, we looked at converting all ID into an internal inner ID.

The second part is to call the algorithm's own estimate () function, which is usually written in a subclass of the corresponding algorithm. Different algorithms correspond to different estimate () methods. At the same time, the return value sometimes contains the content of a details, and the predicted result est is a tuple containing the predicted score and details, which is split here. If the prediction fails, the next function is called: default_prediction (), which we'll introduce right away.

The third part is the content of a clip (), which is to judge whether the predicted results are out of range and carry out a standardization.

Through these three steps, a prediction is completed, and finally a result value predicted by Prediction () is returned.

Default_prediction ()

When there is a problem with the prediction, you choose to call default_prediction (), which calls the global_mean method of trainset itself, and finally returns the average score of the entire data set.

Test () def test (self, testset, verbose=False): predictions = [self.predict (uid, iid, r_ui_trans, verbose=verbose) for (uid, iid, r_ui_trans) in testset] return predictions

The test () function directly calls the previous predict () function for each set of data in testset and returns a list result.

Compute_baselines ()

In this section, the baseline of user and items is calculated only once. If it has already been called in the same dataset, the result will be returned directly next time. The specific interaction between this and the algorithm has not been seen yet, which will be described in more detail later.

Compute_similarities ()

This method builds a similarity matrix, and when the AlgoBase () class initializes, there will be a variable sim_options, which determines what similarity is used to build the similarity matrix.

There is a similarities.cpython-37m-darwin.so file in surprise that encapsulates different similarity calculation methods.

The similarities package built with this file in compute_similarities will eventually return a similarity matrix of the target.

Get_neighbors ()

This method passes in a user ID and an int value k, which returns the ID of k user that is most similar to this user.

The AlgoBase () base class constructs some commonly used methods, which basically include functional interfaces such as fitting, prediction, verification, and so on. These methods are overridden in the specific algorithm, and the latter is called.

After reading the above, have you mastered the method of functional analysis of Algobase () class? If you want to learn more skills or want to know more about it, you are welcome to follow the industry information channel, thank you for reading!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.