Tensorflow

With technology becoming more and more advanced the options seems endless when thinking about technology to put in an app. One technology that is growing more popular is machine learning. Wikipedia describes machine learning as “computer algorithms that improve automatically through experience.” If you want to integrate your application with machine learning, I suggest you first take a look at TensorFlow.

Why my you want machine learning in an app?

Some examples of tensor flow use include a lip syncing app that can tell how closely you can sing the words to a song, image recognition, and even a scavenger hunt app using tensorflow’s machine learning. The possibilities for machine learning is endless

What is Tensorflow?

TensorFlow is an open-source library for machine learning. Machine learning means that through algorithms, apps can be taught how to interpret data. for example airbnb uses tensorflow to teach its app to group photos of a house by the room where the pictures were taken. Twitter uses tensor flow to rank tweets for users based on their likes and follows.

Benefits of TensorFlow

Pre Trained Models: TensorFlow comes with plenty of pre-trained model that may be able to handle the exact type of machine learning that you need. For example if you need a machine learning model for image detection, TensorFlow has plenty of premade models that will handle that for you That way you wouldn't even have to build any models yourself and you could use one of theirs. Some of they premade models include, image object detection, language text detection, audio command detection, and many more.

Characteristics of TensorFLow

TensorFlow comes with a lot of pre-made function that can halp you visualize your data. For example in the code below, we can use the scatterplot function to create a scatter graph of car data.

Other datavisulization that tensorflow has include, charts such a bar charts and many more.

Using the data we created the scatter graph with, we can train a model to predict the miles per gallon of a car based off that car’s horsepower. first we must create the model and tell the model which function to run.

The line of code on line 3 instantiates a tf.model object. This model is sequential because its inputs flow straight down to its output. then on line 6 a dense layer is a type of layer that multiplies its inputs by a matrix (called weights) and then adds a number (called the bias) to the result.

From there you can convert the data into tensor, train the data with your model, and finally you'll be able to make predictions on your data.

TensorFlow can be a complicated but very rewarding way to make your app unique and cutting edge. We are just scratching the surface of what machine learning can do and tensorFlow can be a great jumping off point to start learning machine learning.