Coding Interview Companion for AI and Deep Learning

Explore why grooming skills in AI, data science, and machine learning matter, with an emphasis on image processing and convolutional neural networks, and the strong job market driving demand.

Explore how mean, median, and mode are calculated in Python using numpy and the statistics library, with an example input array and corresponding function calls.

Explore supervised learning by training a model on labeled data with features such as symboling, normalized-losses, make, fuel-type, aspiration, and doors to classify regions and predict new data with accuracy.

Explore the differences between supervised and unsupervised learning, including labelled vs unlabelled data, classification, regression, and clustering. Analyze evaluation methods and how learning is affected by controlled versus pattern-driven environments.

Explore linear regression concepts by modeling salary with years of experience, using a best-fit line y = θ0 + θ1 x to predict pay and understand loss minimization.

Learn to compute euclidean distance in python for image pixels and facial recognition, using the pythagorean (l2-norm) metric with a sample (3,4) to (8,9) giving 7.07.

Explore classification as a supervised learning task, build a from-scratch k-nearest neighbors classifier in Python using Euclidean distance, training data, and majority-vote for unseen points.

Explore a stack data structure and its filo order and top operations through a Python implementation. Review stackNode setup, push and pop operations, and methods like peek, displayStack, and search.

Create a custom stack from scratch, push values 11–19, display the stack, and use foundInStack() to verify 15 is present, then eject five elements and recheck 15 is not found.

Explore visualizing neural networks with two layers and a linear activation, adjusting a dataset of four groups, features, hidden layers, neurons, learning rate, and sigmoid activation.

Learn polar coordinates, where radius r and angle theta define a point, and convert to cartesian x = r cos theta, y = r sin theta in Python NumPy walkthrough.

Learn orientation computation for image processing by implementing a Python point class and a geometry class, using orientation(p1, p2, p3) to classify triplets as clockwise, counter clockwise, or collinear.

Learn to compute a convex hull with the Graham scan: sort points by polar angle, discard collinear interior points, and use a stack to build the hull.

Explore a Python implementation of the convex hull by the Grahamstown algorithm, including stack operations, sorting points, polar angles, and orientation checks to construct the hull.

Explore building a directed graph in Python with an adjacency list using AdjNode and addEdge. Compare adjacency matrix and adjacency list representations for image processing.

Explore how the flood-fill algorithm detects paths and regions in images using depth-first traversal, neighbor definitions, and a visited matrix, with practical maze and paint bucket examples.

Explore image segmentation with BFS and region adjacency graphs to divide images into meaningful regions, apply thresholding and edge detection for deep learning.

Apply breadth first search to detect paths to regions and edges in a pixel graph. The walkthrough covers python code, adjacency lists, and networkx for edge detection in deep learning.

Explore convolutional neural networks, including convolution operations with kernels and padding, max-pooling, and the shift from fully connected layers, illustrated by classifying a 28x28 handwritten digit two and edge detection.

Explore convolutional neural networks through a Python walkthrough that demonstrates a 3x3 kernel convolution on a 6x6 image, producing a 4x4 feature map and explaining padding and the sliding window.

Discover natural language processing, teaching computers to understand and manipulate language, with applications like sentiment analysis, topic modelling, text classification, and sentence segmentation, built around the NLTK toolkit.

Understand how raw NLP text is processed from tokenization and cleaning to vectorization, with cross-validation and spam or ham prediction using NLTK.

Explore natural language processing fundamentals with a python string manipulation warmup. Walk through iterating text, indexing, range traversal, and enumerating to print alternates and separators.

Explore a NLP warmup that demonstrates a primitive text processing function using startswith() and endswith() to find words beginning with app and ending with ion in a sample sentence.

Count words in a sentence using python. Split the text by spaces or periods, then apply len() to get the word count, as in the sample eight words.

Learn how to split a sentence into words with Python's split() in a code walkthrough. See a sample sentence like 'apples and oranges are fruits' being tokenized.

Explore NLP sentence manipulations in Python by reversing a string of words. Tokenize the input into words, reverse the token order with a list comprehension, and join with spaces.

Learn how a Python program reverses each word in a sentence and joins them with spaces. The code walkthrough shows tokenizing the input and reversing tokens with reverseWordSentence().

Explore a naïve paragraph-to-sentence generator by building a mini nltk from scratch in Python, using period as the sentence delimiter, and outlining future needs for exclamations and abbreviations.

learn how to use python's re module to detect words that begin with vowels through a code walkthrough that prints words like apples, oranges, avacado, and eggplant.

Demonstrate a Python regex solution to convert camel-case phrases to spaced lowercase words. Use the re module and findall to extract capitalized segments, lowercase them, and join with spaces.

Develop a custom regular expression from scratch in Python to match vowels at the start of an input string, using a startWithVowel function and a bracketed vowel class.

Demonstrate splitting a paragraph into sentences using Python's regex module. Walk through a code example that imports re, processes a sample paragraph, and prints five sentences like 'NLP is nice'.

Remove punctuation from text using a custom Python function in a mini NLTK-like framework that mimics NLTK from scratch, using a list comprehension and join to produce cleaned text.

Break a paragraph into words with a Python add_spaces function that detects sentence ends, then split the text into tokens.

Explore NLP text pre-processing by walking through a Python tokenization workflow: remove punctuation with string and a list comprehension, convert to lower case, and generate tokens using re-based splitting.

Implement a from-scratch stemming technique using a defined suffix list to trim words to their base forms and compare with the porter stemmer in NLTK for AI text processing.

Explore nlp text pre-processing with a Python custom stemmer that trims words to their roots by removing common suffices, using a cloud of words like watches and eating via regex.

Explore stop words such as the, is, and, and learn to remove them with a Python function using tokenization and a list comprehension, producing words like grapes, sweet, sour.

Learn to create bigrams from tokenized text and use a Python code walkthrough to print consecutive word pairs, building a frequency distribution for vectorization and co-occurrence insights.

Demonstrates bigram generation with list comprehension on token lists, enumerating current and next indices to form bigrams from the example text 'i love programming machine learning with natural language processing'.

Demonstrates how to generate trigram sequences from text, explaining tokens, n-2 trigrams, and a Python implementation with examples like May is month of pink flowers and cool fruits.

Walk through NLP text pre-processing with Python, using regular expressions to split sentences, tokenize text, build trigrams, and identify the most frequent trigram, I like to.

Build a spell checker using a Trie in Python, with 26 child nodes and an ends flag. Implement insert and lookup to verify words from the English dictionary.

Explore an NLP dictionary code walkthrough that uses regular expressions to filter the English lexicon, read a dictionary file line by line, and match words in Python.

Compute the minimum edit distance between strings using dynamic programming, as shown by transforming 'sunday' to 'monday' and implementing a Python editDistance function with a DP table.

Study recursion-based minimum edit distance to measure string similarity, covering insert, delete, substitution costs, and implement a Python editDistance function with examples like amazon to amazing and genome sequences.

Compute the Levenshtein minimum edit distance via dynamic programming with insertion 1, deletion 1, substitution 2. Transforming Sunday to Monday yields distance 4 with two substitutions of su to mo.

Learn how natural language processing vectorizes text by cleaning and tokenizing data, then encodes it into numeric feature vectors using methods like count vectorizer, tf-idf vectorization, and n-grams.

Explore the bag-of-words model in natural language processing, illustrating a document-term matrix built from word counts after preprocessing with punctuation removal and tokenization, using a Python walkthrough.

Explore when to remove or retain stop words in NLP preprocessing, highlighting BoW trimming for pre-processing and preserving stop words in sentiment analysis and product reviews with bi-grams and tri-grams.

Celebrate completing the coding interview companion for AI and deep learning by applying artificial intelligence concepts from scratch, and gear up for whiteboard interviews with NLP and computer vision practice.