Good machine-learning models require lots of high-quality training data. As mentioned before, you should aim for 100,000+ records to train on for a custom model.

Image recognition, for example, with the Vision API.

Deep learning, remember, is a sub-discipline of machine learning.

It's possible to translate entire sentences in Google Sheets with simple formulas:

googletranslate(B3,"en","de")

• Label detection
• Face detection
• OCR (Optical Character Recognition)
• Explicit Content detection
• Landmark detection
• Logo detection

Often, multiple pre-built models are used together in an ML system. For example, if you didn't have free text comments for user reviews but rather had audio from your customer interactions with your call centers, you could first transcribe the audio to text with the Cloud Speech API. Then, use the Natural Language API for sentiment analysis as you saw before.

The API can identify labels within a video and when they occurred, as well as a confidence level.

Dialogflow is a platform for building natural and rich conversational experiences. It was formerly named API.ai until Google acquired it and rebranded it into Dialogflow. At its core, Dialogflow is a powerful natural language understanding engine that processes and understands natural language input. It lets you easily achieve a conversational user experience by handling the natural language understanding for you.

Dialogflow incorporates Google's ever-growing AI expertise, including:

• Machine learning expertise
• Search capabilities
• Speech recognition
• Natural language understanding

Codeless model building with Cloud AutoML.

Train a data model in a custom granularity. The first compute hour is free.

Precision is the number of photos correctly classified as a particular label divided by the total number of photos classified with that label.

Recall is the number of photos classified as a particular label divided by the total number of photos with that label.

Recall that with AutoML Vision, you are primarily doing classification on image data that you provide to the service. AutoML requires no coding experience, while the Vision API requires you to be familiar with invoking APIs using a programming language of your choice.

Once the training data is in Cloud Storage, you need a way for AutoML Vision to access it. You'll create a CSV file where each row contains a URL to a training image and the associated label for that image. This CSV file has been created for you; you just need to update it with your bucket name.

• Copy the template file to your Cloud Shell instance
• Update the CSV with the files in your project
• Upload this file to your Cloud Storage bucket
• Show the bucket to confirm the data.csv file is present

gsutil cp gs://cloud-training/automl-lab-clouds/data.csv .
head --lines=10 data.csv
sed -i -e "s/placeholder/$DEVSHELL_PROJECT_ID-vcm/g" ./data.csv
head --lines=10 data.csv
gsutil cp ./data.csv gs://$DEVSHELL_PROJECT_ID-vcm/
gsutil ls gs://$DEVSHELL_PROJECT_ID-vcm/

A model in BigQuery ML represents what an ML system has learned from the training data.

• Linear regression for forecasting; e.g., predicting sales on a given day.
• Binary logistic regression for classification; e.g., determining if a customer will make a purchase.
• Multiclass logistic regression for classification; e.g., predicting if an input is "low-value," "medium-value," or "high-value."
• K-means clustering for data segmentation; e.g., identifying customer segments.
• Matrix Factorization for creating product recommendation systems.
• Time-series for performing time-series forecasts.
• Boosted Tree for creating XGBoost based classification and regression models.
• Deep Neural Network (DNN) for TensorFlow based classification and regression models.
• AutoML Tables to create best-in-class models without feature engineering or model selection.
• TensorFlow model importing for prediction in BigQuery ML.

BigQuery ML has the following advantages over other approaches to using ML with a cloud-based data warehouse:

• Empowers data analysts to build and run models using existing business intelligence tools and spreadsheets.
• No need to program an ML solution using Python or Java; models are trained using SQL.
• Speeds up model development by bringing ML to the data, rather than exporting and reformatting it.

Exporting data from the warehouse has several disadvantages:

• Increases complexity due to the need for multiple tools.
• Slows down the process due to the time required for data transfer and formatting.
• Requires multiple steps to export data, restricting data experimentation.
• Can be prevented by legal restrictions such as HIPAA guidelines.