When you define a variable, you want it to exist within some boundaries. E.g. a result variable makes sense to exist within a calculate() function, as an internal detail. Outside of the calculate(), the result variable is useless.

The accessibility of variables is managed by scope. You are free to access the variable defined within its scope. But outside of that scope, the variable is inaccessible.

In JavaScript, a scope is created by a function or a code block.

Closure

A closure is the combination of a function bundled together (enclosed) with references to its surrounding state (the lexical environment). In other words, a closure gives you access to an outer function’s scope from an inner function. In JavaScript, closures are created every time a function is created, at function creation time.

A tricky question asked in interviews

In the first loop, declaring var i in the loop is the same as declaring it before the loop, making it a global variable for the loop and its inner functions. on the other hand, let i makes it a local variable for the loop and its inner functions.

The setTimeout function executes the log function 100ms after it’s being called, which is less than the time needed for the loop to be executed.
By the time the first console.log is executed, i has already incremented to 3 and it being a global variable for the loop, all the log functions get i =3. Since i is a local variable for the 2nd loop, the log function will get the value of i from when it was called instead of the value i holds while it’s being executed.

Language : Python
IDE : Google Colab
Code : Github Link

ImageDataBunch:

Bind train, valid and test dataloader in a data object.

• From_folder

Create from imagenet style dataset in path with train,test subfolders. 20% of training data was used for validation (valid_pct=0.2).

path\
  train\
    clas1\
    clas2\
    …
  valid\
    clas1\
    clas2\
    …
  Test\

• get_transforms
  get_transforms(do_flip:bool=True, flip_vert:bool=False, max_rotate:float=10.0, max_zoom:float=1.1, max_lighting:float=0.2, max_warp:float=0.2, p_affine:float=0.75, p_lighting:float=0.75, xtra_tfms:Optional[Collection[Transform]]=None) → Collection[Transform]
  
  Utility func to easily create a list of flip, rotate, zoom, warp, lighting transforms.

• do_flip: if True, a random flip is applied with probability 0.5
• max_rotate: if not None, a random rotation between -max_rotate and max_rotate degrees is applied with probability p_affine
• max_zoom: if not 1. or less, a random zoom between 1. and max_zoom is applied with probability p_affine
• max_lighting: if not None, a random lightning and contrast change controlled by max_lighting is applied with probability p_lighting
• max_warp: if not None, a random symmetric warp of magnitude between -max_warp and maw_warp is applied with probability p_affine
• p_affine: the probability that each affine transform and symmetric warp is applied
• p_lighting: the probability that each lighting transform is applied

Randomly Transforms some images of the dataset. Includes flipping, rotating, zooming, lighting.

• Normalize

This is done to ensure that the images are easy to do mathematical calculations that we are looking for after that. This includes changing the range of values of RGB from 0-255 to -1 to 1.This is because we have 3 color channels namely Red, Green and Blue. For the pixel values we might have some color channels that varies slightly and some that doesn’t. So, we need to normalise the images with mean as 0 and standard deviation as 1.

One another thing to note is that we are using the residual network, which is pretrained. So, we must use the same normalisation that the residual network is using in order to use the best of the pretrained model.

Learn:

• Cnn learner

We used transfer learning since our dataset is small. For our pretrained model ResNet50 was used.

architecture : ResNet50.

Resnet50 has a matrix with 1k weights at the end which is for imagenet dataset. But we don’t need that since our dataset is different, so we replace that matrix with 2 matrices with Relu in between. The last matrix containing weights corresponds to the number of classes we have. We initialize the head randomly and then train the head while freezing the body weights since the body is already pre trained with imagenet dataset and good at extracting features. Then we fine tune the model by unfreezing the backbone layers and training all the layers’ weights with gradually descending learning rates.

Residual block of our ResNet50

Activation : Rectified Linear Unit (ReLU)

ReLU converges faster and is cheap to compute. Also it makes our network sparse (no need to activate all the neurons at once)

• unfreeze

After unfreezing all weights from frozen layers of the model can get updated from their pre-trained state according to loss function.

Frozen by defualt, which is equivalent to using the pre-trained convolution network backbone as a feature extractor,only training the fully connected layer learning classification

• Fit_one_cycle

This Callback allows us to easily train a network using Leslie Smith’s one cycle policy. The original paper has 3 phases but unpublished work has shown even better results by using only two phases :

• Phase 1: We progressively increase our learning rate from lr_max/div_factor to lr_max while progressively decreasing momentum from maximum momentum to minimum, same as the original paper.
• Phase 2: Learning rates follow a cosine annealing from lr_max to 0, as the momentum goes from maximum to minimum with the same annealing.

Learning rate vs iterations momentum vs iterations

This gives a better accuracy than just running the whole cycle with a fixed learning rate.

• max_lr

maintains a low learning rate for initial layers as they need lesser tuning and gradually increase learning rate for later layers which need higher tuning especially fully connected ones. Slice defines the learning rate range. Discriminative learning.

Why do this?

• Because it gives new readers context. What are you about? Why should they read your blog?
• Because it will help you focus your own ideas about your blog and what you’d like to do with it.

The post can be short or long, a personal intro to your life or a bloggy mission statement, a manifesto for the future or a simple outline of your the types of things you hope to publish.

To help you get started, here are a few questions:

• Why are you blogging publicly, rather than keeping a personal journal?
• What topics do you think you’ll write about?
• Who would you love to connect with via your blog?
• If you blog successfully throughout the next year, what would you hope to have accomplished?

You’re not locked into any of this; one of the wonderful things about blogs is how they constantly evolve as we learn, grow, and interact with one another — but it’s good to know where and why you started, and articulating your goals may just give you a few other post ideas.

Can’t think how to get started? Just write the first thing that pops into your head. Anne Lamott, author of a book on writing we love, says that you need to give yourself permission to write a “crappy first draft”. Anne makes a great point — just start writing, and worry about editing it later.

When you’re ready to publish, give your post three to five tags that describe your blog’s focus — writing, photography, fiction, parenting, food, cars, movies, sports, whatever. These tags will help others who care about your topics find you in the Reader. Make sure one of the tags is “zerotohero,” so other new bloggers can find you, too.

Language : C#
IDE : Visual Studio Community 2017
Code : Github Link

What can you do?

• Ask for help on solving questions
• Help solve other’s problems
• Earn points by asking relevant questions and solving answers
• Get tagged as a teacher in our platform after earning a set ammount of points

Language : PHP, HTML, CSS
IDE : Visual Studio Community 2017
Code : Github Link

What can you do?

• Ask for help on solving questions
• Help solve other’s problems
• Earn points by asking relevant questions and solving answers
• Get tagged as a teacher in our platform after earning a set ammount of points