SageMaker Built-in Algorithms

Question 1

Blazing Text

Answer 1

Optimised implementations of Word2Vec and text classification algorithms.

• Word2Vec mode maps words to distributed vectors. Words that are semantically similar correspond to vectors that are close together, therefore word embeddings capture semantic relationships between words. Is useful for downstream NLP tasks
• Text classification mode is for applications that perform web searches, information retrieval, document classification

Question 2

Object2Vec

Answer 2

• General-purpose neural embedding algorithm (i.e. similar to Word2Vec, but generalised for more complex objects such as sentences and paragraphs)
• Can learn low-dimensional dense embeddings of high-dimensional objects, that preserve the semantics of the relationship between pairs of objects
• E.g. sentence-sentence pairs (sentence about soccer and sentence about sport), customer-customer pairs (customer ids for Jane and Jackie), product-product pairs (basketball and football product ids)

Question 3

DeepAR Forecasting

Answer 3

• A supervised learning algorithm for forecasting scalar (one-dimensional) time series using RNNs
• E.g you might have time series groupings for demand for different products, server loads, and requests for webpages. For this type of application, you can benefit from training a single model jointly over all of the time series
• When your dataset contains hundreds of related time series, DeepAR outperforms the standard ARIMA and Exponential Smoothing (ETS) methods.

Question 4

Factorisation Machine

Answer 4

• General-purpose supervised learning algorithm that you can use for both classification and regression tasks. - It is an extension of a linear model that is designed to capture interactions between features within high dimensional sparse datasets economically.
• Factorization machines are a good choice for tasks dealing with high dimensional sparse datasets, such as click prediction and item recommendation.

Question 5

Image Classification Algorithm

Answer 5

• Supervised learning algorithm that supports multi-label classification. It takes an image as input and outputs one or more labels assigned to that image.
• It uses a convolutional neural network (ResNet) that can be trained from scratch or trained using transfer learning when a large number of training images are not available.

Question 6

IP Insights Algorithm

Answer 6

• Unsupervised learning algorithm that learns the usage patterns for IPv4 addresses. It is designed to capture associations between IPv4 addresses and various entities, such as user IDs or account numbers.
• You can use it to identify a user attempting to log into a web service from an anomalous IP address, for example.

Question 7

K-means Algorithm

Answer 7

• SageMaker uses a modified version of the web-scale k-means clustering algorithm. Compared with the original version of the algorithm, the version used by SageMaker is more accurate.
• The algorithm groups observations with similar attribute values

Question 8

K-Nearest Neighbors (k-NN) Algorithm

Answer 8

• Index-based algorithm
• For classification problems, the algorithm queries the k points that are closest to the sample point and returns the most frequently used label of their class as the predicted label.
• For regression problems, the algorithm queries the k closest points to the sample point and returns the average of their feature values as the predicted value.

Question 9

Linear Learner Algorithm

Answer 9

• With the SageMaker LL algorithm, you can simultaneously explore different training objectives and choose the best solution from a validation set.
• You can also explore a large number of models and choose the best. The best model optimizes either of the following:
  
  1. Continuous objectives, such as mean square error, cross entropy loss, absolute error.
  2. Discrete objectives suited for classification, such as F1 measure, precision, recall, or accuracy.
• SageMaker linear learner algorithm provides a significant increase in speed over naive hyperparameter optimization techniques

Question 10

Neural Topic Model (NTM) Algorithm

Answer 10

• Unsupervised learning algorithm that is used to organize a corpus of documents into topics that contain word groupings based on their statistical distribution.
• Documents that contain frequent occurrences of words such as “bike”, “car”, “train”, “mileage”, and “speed” are likely to share a topic on “transportation” for example.
• Because the method is unsupervised, only the number of topics, not the topics themselves, are prespecified.

Question 11

Object Detection Algorithm

Answer 11

• Detects and classifies objects in images using a single deep neural network.
• It is a supervised learning algorithm that takes images as input and identifies all instances of objects within the image scene.
• The object is categorized into one of the classes in a specified collection with a confidence score that it belongs to the class. Its location and scale in the image are indicated by a rectangular bounding box

Question 12

Semantic Segmentation Algorithm

Answer 12

• Provides a fine-grained, pixel-level approach to developing computer vision applications. It tags every pixel in an image with a class label from a predefined set of classes.
• Tagging is fundamental for understanding scenes, which is critical to an increasing number of computer vision applications, such as self-driving vehicles, medical imaging diagnostics, and robot sensing.
• Whereas the Image Classification Algo classifies whole images and the Object Detection Algo detects and classifies objects, the Semantic Segmentation Algo classified every single pixel, therefore providing information about shapes of objects

Question 13

Sequence-to-Sequence Algorithm

Answer 13

• Supervised learning algorithm where the input is a sequence of tokens (for example, text, audio) and the output generated is another sequence of tokens.
• Example applications include: machine translation, text summarization (input a longer string of words and predict a shorter string of words that is a summary), speech-to-text.

Question 14

XGBoost Algorithm

Answer 14

• Popular and efficient open-source implementation of the gradient boosted trees algorithm.
• Supervised learning algorithm that attempts to accurately predict a target variable by combining an ensemble of estimates from a set of simpler, weaker models.
• Robustly handles a variety of data types, relationships, and distributions, and because of the large number of hyperparameters that can be tweaked and tuned for improved fits