Deep Learning with TensorFlow

TensorFlow is a primary software tool of deep learning, as a data science you must be aware about TensorFlow. Checkout these questions if you are looking for a job in data science.

Q.1 Define Image Dashboard in TensorBoard.
The Image Dashboard is used for displaying png files that were saved via a tf.summary.image. The dashboard is configured in such a way so that each row corresponds to a different tag, and each column corresponds to a run. The image dashboard also supports arbitrary pngs which can be used for embedding custom visualizations into TensorBoard.
Q.2 What is Audio Dashboard?
The Audio Dashboard is used for embedding playable audio widgets for audio stored via a tf.summary.audio. The dashboard is configured so that each row corresponds to a different tag, and each column corresponds to a run. The Audio dashboard always embeds the latest audio for each tag.
Q.3 Describe the Graph Explorer in TensorFlow.
The Graph Explorer can be used while visualizing a TensorBoard graph. It is also responsible for allowing inspection of the TensorFlow model. However, for getting the best use of the graph visualizer, one should use name scopes to group the ops in a graph hierarchically. Otherwise, the graph may be difficult to decipher.
Q.4 Define an Embedding Projector.
The Embedding Projector enables visualizing high-dimensional data. For example, one can view input data after it has been embedded in a high-dimensional space by model. The embedding projector reads data from the model checkpoint file. It may be configured with additional metadata, like a vocabulary file or sprite images.
Q.5 Is it possible to run TensorBoard without TensorFlow?
If you do not have TensorFlow installed then, TensorBoard 1.14+ can be run but with a low feature set. The limitation is that as of TensorFlow 1.14, only the following plugins are supported like scalars, custom scalars, image, audio, graph, projector (partial), distributions, histograms, text, PR curves, mesh. Also, there is no support for log directories on Google Cloud Storage.
Q.6 How can we report a vulnerability in TensorFlow?
For reporting about any security issue directly, email at [email protected]. The report to this email is delivered to the security team at TensorFlow. The emails are then acknowledged within 24 hours, and a detailed response is provided within a week along with the next steps.
Q.7 Name the components used for deploying a lite model file in TensorFlow.
1. Java API Java API is a wrapper around C++ API on Android. 2. C++ API C++ API loads the TensorFlow Lite model and calls the interpreter. 3. Interpreter The interpreter can be used for executing the model. It uses particular kernel loading, which is a unique feature of TensorFlow Lite.
Q.8 Where can we use word embedding? Name two models used in word embedding.
Word embedding is used in Natural Language Processing as a representation of words. Moreover, they can be used in TensorFlow where it is also known as Word2vec. However, the two models used in word embedding are Firstly, the continuous bag of words model Secondly, the skip-gram model.
Q.9 What are TensorFlow Estimators?
Estimators is a high-level API that reduces much of the code you previously required to write while training a TensorFlow model. Estimators are very flexible, and it enables overriding of the default behavior if there is any specific requirement for the model.
Q.10 Explain the ways of creating a model using estimators.
There are two possible ways of creating a model using Estimators: 1. Pre-made Estimator These are predefined estimators that are created to generate a specific type of model. For example, DNNClassifier, which is a pre-made estimator. 2. Estimator (base class) It provides complete control of how a model should be created by using a model_fn function. However, these functions are consumed by the class tf.estimator.Estimator. The functions return an initialized estimator, upon which we can call .train, .eval, and .predict.
Q.11 Is it possible to deploy a TensorFlow in container software?
Tensorflow can also be used with containerization tools such as Docker. And, it could be used to deploy a sentiment analysis model which uses character-level ConvNet networks for text classification.
Q.12 What are Neural Networks
Neural networks refer to a network of elemental processing entities that together make a complex form. There can be Artificial Neural Networks and Biological Neural Networks. The use of artificial neural networks is more common as they try to imitate the mechanics of the human brain.
Q.13 Name dome statistical distribution functions provided by TensorFlow.
A large variety of statistical distributions is available which is provided by TensorFlow and located inside tf.contrib.distributions. This consists of distributions like Beta, Bernoulli, Chi2, Dirichlet, Gamma, Uniform, etc. These are important building blocks when it comes to building machine learning algorithms, especially for probabilistic approaches like Bayesian models.
Q.14 What do you understand by Tensorflow?
TensorFlow is defined as a machine learning library created by the Brain Team of Google and made open source in 2015. Tensorflow can be defined as is a low-level toolkit for doing complicated math and it offers the users customizability to build experimental learning architectures, to work with them and to turn them into running software.
Q.15 Under what situation will you find overfit condition of your model in TensorFlow?
Indeed there are variations in the training data or data that needs to be verified through TensorFlow. Such that if the variations are very large in the data, probably it could lead to this problem. One of the best possible solution is to remove the noise from the available data upto the possible extent.
Q.16 What is Bias-Variance decomposition?
Bias-Variance decomposition is generally used to decompose problems such as errors that occur during learning in different algorithms. Such that bias keeps reducing if the data is to be made complex. In which case trading off the variance and bias are very essential to get results that are totally free from errors.
Q.17 What do you understand by discriminative and generative method in TensorFlow?
Discriminative Method - All the distinction among the different types of data can simply be learned with the help of discriminative method. Generative model - This method is used for understanding a specific format of same. The tasks that can also be handled with both these approaches need to be classified in a well-defined order first.
Q.18 What are the advantages of TensorFlow?
Some of the advantages of TensorFlow are -
1. Platform flexibility
2. Easily trainable on CPU as well as GPU for distributed computing.
3. TensorFlow has auto differentiation capabilities
4. Advanced support for threads, asynchronous computation, and queue es.
5. Customizable and open source.
Q.19 How would you define TensorFlow loaders?
Tensorflow Loaders are primarily used for adding algorithms and data backends one of which is tensorflow itself. For instance a loader can be implemented to load, access and unload a new type of servable machine learning model.
Q.20 What are Tensors?
Tensors are higher dimensional arrays which are used to represent a multitude of data in the form of numbers in the computer programming. There are other n-d array libraries available like Numpy, while the TensorFlow stands apart from them as it offers ways to create tensor functions and automatically compute derivatives.
Q.21 What is a TensorBoard?
TensorBoard, a suit of visualizing tools, is a simple solution to Tensorflow offered by the creators that lets you plot quantitative metrics, and visualize the graphs about the graph with additional data like images to pass through it.
Q.22 List a few advantages of TensorFlow?
It has platform flexibility
It is easily docile on CPU as well as GPU for distributed computing.
TensorFlow has auto differentiation capabilities
It has an advanced support for queue es, threads, and asynchronous computation.
It is a customizable and open source.
Q.23 List a few limitations of Tensorflow.
If imported in the same scope it has GPU memory conflicts with Theano.
No support for OpenCL
Knowledge of advanced calculus and linear algebra is required along with a pretty good understanding of machine learning.
Q.24 What are TensorFlow servable?
These are the central rudimentary units in TensorFlow Serving. Servable are the objects that clients use, to carry-out the computation.
The size of a servable is flexible. A single servable might involve anything from a lookup table to a single model to a tuple of inference models.
Q.25 What do the TensorFlow managers do?
Tensorflow Managers handle the full lifecycle of Servable, including:
Loading Servable
Serving Servable
Unloading Servable
Q.26 What are TensorFlow loaders?
Tensorflow Loaders are utilized for the addition of algorithms and data backends one of which is tensorflow itself. For instance: a loader can be implemented to load, access and unload a new type of servable machine learning model.
Q.27 What do you mean by sources in TensorFlow?
Sources are in simple terms, modules that find and provide servable. Each Source provides zero or more servable streams. One Loader is supplied for each servable version it makes accessible to be loaded.
Q.28 How does TensorFlow utilizes the python API?
Python is the most recognisable and “the main” language when it comes to TensorFlow and its development. It's the first language supported by TensorFlow and still supports most of the features. It seems as TensorFlow’s functionality first define in Python and then moved to C++.
Q.29 What are the general advantages of using the Artificial Neural Networks?
They use to find answers to complex problems in a stepwise manner. All the information received by a network can easily be in any format. Along with a good tolerance capability they also make use of real-time operations.
Q.30 They use to find answers to complex problems in a stepwise manner. All the information received by a network can easily be in any format. Along with a good tolerance capability they also make use of real-time operations.
The another name of Recall is the true positive rate. It is the total figure of costiveness a model can generally claim. The predictive value which is predominantly positive in nature is Precision. The disc-similarity between the true positive rate and claimed positive rate can be defined with the help of both these options.
Q.31 Name some products built using TensorFlow?
TensorFlow built the following products:
Teachable Machine
Synch
Handwriting
Giorgio Camthat
Q.32 What advantages TensorFlow has over other libraries?
Debugging facility, scalability, visualization of data, pipelining and many more.
Q.33 How can you assure that overfitting situation is not arriving with a model you are using?
Users need to make sure that their model isn't complex and not have only simple statement. Variance takes into the account and the noise deletes from the model data. Techniques like k-fold and LASSO can also help.
Q.34 What exactly do you know about a ROC curve and its functioning?
ROC or region of convergence used to reflect data rates which classify as true positive and false positive. Presented in the form of graphs, it can use as a proximity to swap operations related to different algorithms.
Q.35 In the machine learning context, state how reliable and useful Bayes’ theorem is?
Baye’s theorem is useful for determining the probability of an event, obtained by dividing the actual positive rate by the false positive rate. Some of the very complex questions and challenges can solve using a simple approach and eliminated with the help of this theorem.
Q.36 What differences do Type I and Type II errors hold?
Type I error means a false positive value, whereas Type II error means a false negative.
Q.37 Give a strategy to handle a situation indicating an imbalanced dataset?
This usually happens when a vast set of data keep in a single class. One of the possible solutions is sampling the dataset again and the other one being the migration of data to parallel classes. While the dataset should not be damaged.
Q.38 Name the types of Tensors.
There are three types of Tensors used for creating neural network models: 1. Constant Tensor Constant Tensors are used as constants, as the name suggests. They create a node that takes a value and does not change it. A constant can be created using tf.constant. 2. Variable Tensor Variable Tensors are the nodes that provide their current value as output. It means that they can retain their value over multiple executions of a graph. 3. Place Holder Tensor These are used for assigning data at a later time. Placeholders are the nodes whose value is fed at the time of execution.
Q.39 List the client languages that are supported in TensorFlow.
TensorFlow provides support for multiple client languages, one of the best among them is Python. There are some experimental interfaces that are available for C++, Java, and Go. Language bindings for many other languages such as C#, Julia, Ruby, and Scala are created and supported by the open-source community.
Q.40 Name the three working components of TensorFlow Architecture.
TensorFlow architecture works in three parts: Preprocessing the data Building the model Training and estimating the model
Q.41 What are the ways for loading data into TensorFlow?
There are two ways to load the data: 1. Load data in memory This s the easiest method. All the data is loaded into memory as a single array. One can write Python code that is unrelated to TensorFlow. 2. Tensorflow data pipeline TensorFlow has built-in APIs which help in loading the data, performing the operations, and feeding the machine learning algorithm easily. This method is mostly used for large datasets.
Q.42 Can you provide the basic steps for TensorFlow algorithms?
Firstly, importing data, generating data, or setting a data pipeline through placeholders. Secondly, feeding the data through the computational graph. Thirdly, evaluating output on the loss function. Then, using backpropagation to modify the variables. Lastly, repeating until stopping condition.
Q.43 Describe the use cases of TensorFlow?
Tensorflow is an essential tool for deep learning. This has five use cases, they are: Text-Based Applications Voice/Sound Recognition Time Series Image Recognition Video Detection
Q.44 Name the platforms where we can run a TensorFlow.
TensorFlow can run on different platforms like: Operating Systems such as Windows, OS, and Linux Cloud Web Service Mobile OS like IOS and Android
Q.45 Explain the benefits of TensorFlow over other libraries.
Firstly, TensorFlow provides easily scaled machine learning applications and infrastructure. Secondly, visualizing the graph is very straightforward in TensorFlow. TensorBoard(a suite of visualization tools) is used for visualizing TensorFlow graphs. Thirdly, tfdbg is a specialized debugger for TensorFlow. It allows us to view the internal structure and states of running TensorFlow graphs during training and inference. Lastly, TensorFlow's Dataset module tf.data is used for creating efficient pipelines for images and text.
Q.46 Define the term Deep Speech.
DeepSpeech refers to an open-source engine used for converting Speech into Text. It uses a model which is trained by machine learning techniques. It is based on Baidu's Deep Speech research paper and uses Google's TensorFlow for making the implementation easier. Syntax: ./deepspeech.py
Q.47 Explain TensorFlow abstractions.
TensorFlow contains abstraction libraries such as TF-Slim and kereas, which provide simplified high-level access to TensorFlow. Such abstractions help in streamlining the construction of data flow graphs. Further, TensorFlow abstractions also help to make the code cleaner and reduce the length of codes drastically.
Q.48 Explain the APIs used inside the TensorFlow project?
APIs inside TensorFlow are dependent on the Python language. They have low-level options for the users such as tf.manual or tf.nn.relu used for build Neural Network Architecture. These APIs are also used for designing a deep neural network having higher levels of abstraction.
Q.49 Name the APIs used outside the TensorFlow project?
There are a few APIs used outside the TensorFlow project, which are: 1. TFLearn TFLearn provides a high-level API which makes neural network building and training fast and easy. Its API can be indicated as tf.contrib.learn. 2. TensorLayer TensorLayer is a TensorFlow-based deep learning and reinforcement learning library. It is designed for researchers and engineers. It provides an extensive collection of customizable neural layers/functions which are crucial for building real-world AI applications. 3. Sonnet Sonnet is a library that is built on top of TensorFlow for creating complex neural networks. It is a part of Google's DeepMind project, which features a modular approach.
Q.50 Explain the process for creating tensors from Python objects.
We can create tensors such as NumPy arrays and lists with the help of Python objects. We can easily perform it using tf.convert_to_tensor() operation.
Q.51 Define variables in TensorFlow.
Variables in TensorFlow are also known as tensor objects which hold the values that can be modified during the execution of the program. Further, it is used to represent a shared, persistent state manipulated by the program.
Q.52 Define Scalar Dashboard.
Scaler Dashboard visualizes scaler statistics that vary over time. It uses a simple API for executing such visualizations. For example, We might want to assess the model's loss or learning rate.
Q.53 Define Histogram Dashboard.
The Histogram Dashboard is used for displaying how the statistical distribution of a Tensor varies over time. It helps in visualizing the data recorded via tf.summary.histogram. Each chart shows the temporal "slices" of data, where each slice is a histogram of the tensor at a given step. However, if a Histogram mode is changed from "offset" to "overlay", the perspective will rotate.
Q.54 What is Distribution Dashboard in TensorFlow?
The Distribution Dashboard is used for visualizing histogram data from tf.summary.histogram. It shows some high-level statistics on a distribution. Each line on the chart is used to represent a percentile in the distribution over the data. For example, the bottom line displays how the minimum value changes over time and the line in the middle displays how the median changes. Reading it from top to bottom, the lines provide the following meaning: [maximum, 93%, 84%, 69%, 50%, 31%, 16%, 7%, minimum].
Get Govt. Certified Take Test