DEV Community: Favor Onuoha

Accounting 101: Learn how to build financial applications

Favor Onuoha — Sun, 11 Jan 2026 20:47:31 +0000

Fintech is one of the most interesting domains in tech, yet there's surprisingly little beginner-friendly content for developers looking to get started. At its core, FinTech combines traditional finance and accounting knowledge with software engineering to build scalable financial tools.

You might be wondering: if it's "FinTech" (Financial Technology), why are we learning about accounting instead of finance? Good question. Let me explain.

Finance vs. Accounting

Finance and Accounting are related but different fields.

Accounting is about recording and reporting financial transactions. This includes preparing the three core financial statements: the Income Statement (or Profit & Loss Statement), the Balance Sheet, and the Cash Flow Statement. It gives you a historical view of a company's financial performance at a specific point in time. It's essential for compliance and forms the foundation for financial analysis.

Finance is broader. It involves managing money and making strategic decisions based on accounting data. Finance activities include investing, borrowing, lending, budgeting, and forecasting. While accounting looks backward, finance looks forward - using past data to make decisions about future projects, resource allocation, and profitability.

The key point is that understanding finance requires foundational knowledge of accounting, as accounting data is the basis for financial decision-making.

Bookkeeping vs. Accounting

Another term you'll hear is Bookkeeping. While related to accounting, they serve different purposes.

Bookkeeping	Accounting
Recording financial transactions	Summarizing and analyzing financial data
Posting debits and credits	Preparing financial statements
Reconciling bank statements	Analyzing operational costs
Processing accounts payable/receivable	Completing tax returns
Maintaining ledgers and accounts	Advising on financial decisions

Bookkeeping focuses on maintaining accurate records. Accounting focuses on interpreting those records to produce financial statements and actionable insights.

What Are Financial Models?

A Financial Model is a tool used to forecast a business's financial performance. Models perform calculations and generate recommendations based on that analysis.

Common uses include:

Business valuation
Financial planning and budgeting
Forecasting
Investment analysis
Raising capital

This guide focuses on Accounting, teaching you the fundamentals needed to build software that records and reports financial transactions.

What is Accounting?

Accounting is the systematic process of recording, categorizing, and reporting financial transactions. Its purpose is to provide a clear and accurate picture of an entity's economic activities.

Key Aspects of Accounting

Recording Transactions: The fundamental step where every financial transaction is documented. These records form the basis for everything else.
Classifying Transactions: Transactions are categorized into accounts like sales revenue, expenses, assets, liabilities, and equity.
Preparing Financial Statements: Transaction data is used to prepare the Income Statement, Balance Sheet, and Cash Flow Statement.
Auditing and Compliance: Ensuring records are accurate and comply with legal, regulatory, and tax requirements.
Analysis and Advisory: Analyzing financial data to understand performance, identify trends, and provide recommendations.
Budgeting and Forecasting: Preparing budgets and forecasts to help businesses plan.

Accounting is essential not just for businesses but also for individuals, non-profits, and government organizations.

The Accounting Equation

The foundation of all accounting is this equation:

Assets = Liabilities + Equity

This equation must always balance. Every transaction affects at least two accounts in a way that maintains this equality. This is the basis of double-entry bookkeeping.

Double-Entry Bookkeeping

Double-entry bookkeeping is the standard system where every transaction is recorded in at least two accounts:

One account is debited
Another account is credited
Total debits must always equal total credits

This provides built-in error checking and creates a complete audit trail.

Example: When a company receives $1,000 cash from a customer:

Debit Cash (Asset): $1,000 (increases)
Credit Accounts Receivable (Asset): $1,000 (decreases)

Accounts: The Building Blocks

An account is a record in the General Ledger used to collect and store debit and credit amounts. Following the double-entry principle, every transaction affects at least two accounts.

Types of Accounts

Account Type	Normal Balance	Examples
Assets (Dr.)	Debit	Cash, inventory, property/plant/equipment, patents
Liabilities (Cr.)	Credit	Loans, accounts payable, mortgages, deferred revenue
Equity (Cr.)	Credit	Share capital, retained earnings, additional paid-in capital
Revenue (Cr.)	Credit	Sales revenue, service revenue, interest income
Expenses (Dr.)	Debit	Cost of goods sold, salaries, rent, utilities

How Debits and Credits Affect Accounts

Account Type	Increased by	Decreased by
Assets	Debit	Credit
Liabilities	Credit	Debit
Equity	Credit	Debit
Revenue	Credit	Debit
Expenses	Debit	Credit

These account types appear on their respective sides of the General Ledger (GL) - debits on the left, credits on the right.

Contra Accounts

Contra accounts are special accounts used to offset the balance of a related account. They carry opposite balances to their paired accounts.

Think of them as mechanisms that partially reverse the effects of another account, reducing its reported value on financial statements.

Characteristics:

Opposite Balance: If the related account has a debit balance, the contra has a credit balance (and vice versa)
Paired with Specific Accounts: Always linked to and offsets a specific account
Net Reporting: Displayed as a single net amount on financial statements

Common Contra Accounts

Contra Account	Paired With	Purpose
Accumulated Depreciation	Property, Plant & Equipment	Records total depreciation over time
Allowance for Doubtful Accounts	Accounts Receivable	Estimates uncollectible receivables
Sales Returns and Allowances	Sales Revenue	Records returned goods or post-sale allowances

Example: A company buys a machine for $10,000 that depreciates by $1,000 annually. The accumulated depreciation account will show a credit balance of $1,000 after year one, $2,000 after year two, and so on.

Debits and Credits

Debit (Dr.): An entry on the left side of an account. Increases asset and expense accounts; decreases liability, equity, and revenue accounts.

Credit (Cr.): An entry on the right side of an account. Increases liability, equity, and revenue accounts; decreases asset and expense accounts.

Journal Entries

A journal entry is the record of a financial transaction before it's posted to the General Ledger. They're recorded in chronological order in the General Journal.

Each journal entry includes:

Date of the transaction
Accounts affected (debited and credited)
Amounts
Description explaining the transaction

Example:

The General Ledger (GL)

The General Ledger is the master record of all financial transactions within an organization. It's the central repository for accounting data from all sub-ledgers and journals (sales journal, purchases journal, cash receipts journal, etc.).

A GL contains:

Individual accounts are organized by the Chart of Accounts
Transaction details (date, description, amount)
Debit and credit entries

Chart of Accounts

The Chart of Accounts (COA) is the master list of all accounts used by an organization, organized by category. It provides a standardized numbering system.

Typical structure:

Account Range	Category
1000-1999	Assets
2000-2999	Liabilities
3000-3999	Equity
4000-4999	Revenue
5000-5999	Cost of Goods Sold
6000-6999	Operating Expenses

The COA is one of the first things you'll design when building accounting software.

Trial Balance

A Trial Balance is a report that lists all General Ledger accounts and their balances at a specific point in time. Its purpose is to verify that total debits equal total credits.

If the Trial Balance doesn't balance, there's an error somewhere that must be found and corrected before preparing financial statements.

Fiscal Year and Accounting Periods

A fiscal year is the 12-month period a company uses for accounting and financial reporting. Unlike a calendar year (January-December), a fiscal year can start on any date.

Common examples:

US Federal Government: October 1 - September 30
Many retail companies: February 1 - January 31 (to capture the holiday season)
UK companies: April 1 - March 31

Accounting periods are the intervals at which financial statements are prepared - typically monthly, quarterly, or annually.

Accrual vs. Cash Basis Accounting

Understanding when to recognize revenue and expenses is critical for building payment systems.

Aspect	Cash Basis	Accrual Basis
Revenue recognized	When cash is received	When earned (regardless of payment)
Expenses recognized	When cash is paid	When incurred (regardless of payment)
Complexity	Simple	More complex
Best for	Small businesses	Most FinTech applications
GAAP/IFRS compliant	No	Yes

Example with a Subscription Service:

Customer pays $120 on January 1 for a 12-month subscription
Cash basis: Record $120 revenue in January
Accrual basis: Record $10 revenue each month (earned as service is provided)

For developers, this means your system needs to track both when money moves AND when revenue/expenses are recognized. These are often different events.

Immutable Ledgers and Audit Trails

In FinTech, you never update or delete a financial transaction. This is non-negotiable for regulatory compliance and auditability.

The rules:

Transactions are immutable - once recorded, they cannot be modified
Corrections are reversals - to fix an error, create a reversing entry, then create the correct entry
Everything is timestamped - record who did what, when
Soft deletes only - mark records as deleted, never actually delete them

Reversal Pattern

Instead of updating a wrong entry, you create a reversal:

Original (Wrong):
  Jan 15: Debit Office Supplies $500, Credit Cash $500

Reversal:
  Jan 16: Debit Cash $500, Credit Office Supplies $500

Correcting Entry:
  Jan 16: Debit Office Equipment $500, Credit Cash $500

For developers: your database schema should NOT allow UPDATE or DELETE on transaction tables. Build your APIs to only INSERT.

Common FinTech Transactions

Here's how everyday FinTech operations translate to journal entries:

Customer Deposits Money into Wallet

Debit:  Bank Account (Asset)           $100
Credit: Customer Wallet Liability      $100

The money is in your bank, but you owe it to the customer.

Customer Withdraws Money

Debit:  Customer Wallet Liability      $50
Credit: Bank Account (Asset)           $50

Transfer Between Users (User A to User B)

Debit:  User A Wallet Liability        $25
Credit: User B Wallet Liability        $25

No real money moved - just updated who you owe.

Platform Charges a Transaction Fee

Debit:  Customer Wallet Liability      $1
Credit: Fee Revenue (Revenue)          $1

Customer Requests a Refund

Debit:  Refunds (Contra-Revenue)       $20
Credit: Customer Wallet Liability      $20

Loan Disbursement

Debit:  Loans Receivable (Asset)       $1,000
Credit: Bank Account (Asset)           $1,000

Loan Repayment (Principal + Interest)

Debit:  Bank Account (Asset)           $110
Credit: Loans Receivable (Asset)       $100
Credit: Interest Income (Revenue)      $10

From Theory to Code: Database Schema

Here's a minimal database schema for an accounting system:

Core Tables

-- Chart of Accounts
CREATE TABLE accounts (
    id              SERIAL PRIMARY KEY,
    account_code    VARCHAR(20) UNIQUE NOT NULL,
    account_name    VARCHAR(100) NOT NULL,
    account_type    VARCHAR(20) NOT NULL,  -- ASSET, LIABILITY, EQUITY, REVENUE, EXPENSE
    normal_balance  VARCHAR(10) NOT NULL,  -- DEBIT or CREDIT
    parent_id       INTEGER REFERENCES accounts(id),
    is_active       BOOLEAN DEFAULT true,
    created_at      TIMESTAMP DEFAULT NOW()
);

-- Journal Entries (Header)
CREATE TABLE journal_entries (
    id              SERIAL PRIMARY KEY,
    entry_date      DATE NOT NULL,
    description     TEXT,
    reference_no    VARCHAR(50),
    created_by      INTEGER NOT NULL,
    created_at      TIMESTAMP DEFAULT NOW(),
    -- Immutability: no updated_at column!
    is_reversed     BOOLEAN DEFAULT false,
    reversed_by_id  INTEGER REFERENCES journal_entries(id)
);

-- Journal Entry Lines (Details)
CREATE TABLE journal_entry_lines (
    id              SERIAL PRIMARY KEY,
    journal_entry_id INTEGER REFERENCES journal_entries(id),
    account_id      INTEGER REFERENCES accounts(id),
    debit_amount    DECIMAL(15,2) DEFAULT 0,
    credit_amount   DECIMAL(15,2) DEFAULT 0,
    description     TEXT,

    -- Enforce: only debit OR credit, not both
    CONSTRAINT check_debit_or_credit 
        CHECK (
            (debit_amount > 0 AND credit_amount = 0) OR 
            (credit_amount > 0 AND debit_amount = 0)
        )
);

-- Account Balances (Denormalized for performance)
CREATE TABLE account_balances (
    id              SERIAL PRIMARY KEY,
    account_id      INTEGER REFERENCES accounts(id),
    period_date     DATE NOT NULL,  -- Last day of accounting period
    debit_total     DECIMAL(15,2) DEFAULT 0,
    credit_total    DECIMAL(15,2) DEFAULT 0,
    balance         DECIMAL(15,2) DEFAULT 0,
    UNIQUE(account_id, period_date)
);

Entity Relationship

Trial Balance Query

-- Generate Trial Balance
SELECT 
    a.account_code,
    a.account_name,
    a.account_type,
    SUM(jel.debit_amount) as total_debits,
    SUM(jel.credit_amount) as total_credits,
    SUM(jel.debit_amount) - SUM(jel.credit_amount) as balance
FROM accounts a
LEFT JOIN journal_entry_lines jel ON a.id = jel.account_id
LEFT JOIN journal_entries je ON jel.journal_entry_id = je.id
WHERE je.entry_date <= '2026-01-31'  -- As of date
  AND je.is_reversed = false
GROUP BY a.id, a.account_code, a.account_name, a.account_type
ORDER BY a.account_code;

-- Verify Trial Balance (should return 0)
SELECT 
    SUM(debit_amount) - SUM(credit_amount) as difference
FROM journal_entry_lines jel
JOIN journal_entries je ON jel.journal_entry_id = je.id
WHERE je.is_reversed = false;

The Double-Entry Constraint

Every journal entry must balance. Enforce this at the application level:

function createJournalEntry(entry: JournalEntry): void {
    const totalDebits = entry.lines.reduce((sum, l) => sum + l.debitAmount, 0);
    const totalCredits = entry.lines.reduce((sum, l) => sum + l.creditAmount, 0);

    if (totalDebits !== totalCredits) {
        throw new Error(
            `Entry does not balance: Debits=${totalDebits}, Credits=${totalCredits}`
        );
    }

    // Proceed with insert...
}

Reconciliation

Reconciliation is the process of matching your internal ledger with external statements (bank statements, payment processor reports, etc.).

Why it matters:

Detects discrepancies (fraud, errors, missing transactions)
Required for regulatory compliance
Builds trust in your financial data

Basic Reconciliation Flow

Export transactions from your ledger for a period
Import the bank statement for the same period
Match transactions (by amount, date, reference)
Investigate unmatched items
Create adjusting entries if needed
Document the reconciliation

Build reconciliation tools early. You'll need them for debugging, customer support, and audits.

Key Takeaways

Concept	Summary	Developer Implication
Accounting Equation	Assets = Liabilities + Equity	Your database must always satisfy this
Double-Entry Bookkeeping	Every transaction affects 2+ accounts	Validate that debits = credits before saving
Immutable Ledger	Never update/delete transactions	Design INSERT-only transaction tables
Accrual Accounting	Recognize revenue when earned	Track both cash movement AND recognition dates
Chart of Accounts	Master list of all accounts	Design this schema first - it's your foundation
Trial Balance	Debits must equal credits	Build this as a data integrity health check
Audit Trail	Who did what, when	Log everything with timestamps and user IDs
Reconciliation	Match the internal ledger to external sources	Build tooling for this from day one

What's Next?

You now have the accounting foundation needed to build FinTech software. This is Part 1 of a 6-part series. By the end, you'll have built a functional accounting system.

Part	Topic	What You'll Build
1	Accounting Fundamentals	(this post)
2	Building Your First Ledger	Simple API with accounts and journal entries
3	Accounts Receivable (Invoicing)	Create invoices, mark as paid, and track outstanding
4	Accounts Payable (Bills)	Record bills, track what you owe
5	Bank Reconciliation	Match transactions between the ledger and bank statements
6	Financial Reports	Generate an Income Statement and Balance Sheet from your data

45+ Free Developer Resources for Web Developers (Oversimplified Version)

Favor Onuoha — Tue, 08 Jun 2021 16:22:17 +0000

I have compiled a list of great resources for aspiring and professional developers. Hope you enjoy and be sure to check me out on Twitter, I post similar great content.

Beginner Roadmap (Full-Stack)

1) Learn HTML
2) Learn CSS
3) Learn JavaScript
4) Learn React JS
5) Learn Angular JS
6) Learn Bootstrap JS
7) Learn Node JS
8) Learn Express JS
9) Learn MongoDB
10) Learn Mongoose
11) Learn RESTful Architecture
12) Learn Building APIs with Node JS

NoSQL DBs & RDBMS

13) Learn SQL
14) Learn NoSQL Concepts
15) Learn RDBMS Concept
16) Learn Firebase Realtime Database
17) Learn Firebase Cloud Firestore
18) Learn DynamoDB
19) Learn CouchDB
20) Learn Redis Cache
21) Learn PostgreSQL
22) Learn MySQL
23) Learn MSSQL
24) Learn SQLite
25) Learn Oracle
26) Learn RethinkDB

Server Side Languages & Frameworks

27) Learn Java
28) Learn C#
29) Learn Python
30) Learn Ruby
31) Learn Rust
32) Learn Golang
33) Learn Erlang
34) Learn PHP
35) Learn Django
36) Learn Flask
37) Learn ASP.NET API
38) Learn ASP.NET MVC
39) Learn Ruby on Rails
40) Learn Spring
41) Learn Laravel

Cloud Services & Orchestration

42) Learn GCP
43) Learn AWS
44) Learn Digital Ocean Hosting
45) Learn Linode Hosting
46) Learn Docker
47) Learn Kubenetes

Testing & Version Control

48) Learn Git
49) Learn Jest
50) Learn GitHub

Conclusion

Thank you for reading 🔥🔥🔥, and I hope you found this article useful 🔥

If you have any questions or your just wanna chat, feel free to DM me

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

Let's talk about CSS [Easy Roadmap]

Favor Onuoha — Wed, 02 Jun 2021 15:48:31 +0000

For most people like me, CSS was the bane of my existence. It never listens to me when I give it simple instructions and it acts as if it has a mind of its own, relatable right ?. However, CSS acting the way it did was because I didn't understand it back then.

Understanding how layouts and hierarchy works in CSS is the key to mastering the styling language and that is why I have prepared an easy roadmap for mastering CSS

Let's get straight to it...

1) Syntax

Just like every other language out there, you have to master the CSS syntax to be able to confidently write it.

To master this, you will have to understand basic core CSS principles.

You can find more info on specificity here

2) Hierarchy

Hierarchy is important in CSS. Understanding how the Parent-Child hierarchy works in CSS helps you write better CSS code.

This idea of CSS having a hierarchy and how it organizes and accesses elements is known as Specificity

You can find more info on specificity here

3) Layout

Know your CSS Layouts!, this is very crucial to understanding how to position elements on a webpage.

This is by far the most difficult part of learning CSS and here are some links I have prepared to help you find great resources on CSS Layouts

Creating great layouts is your key to mastering CSS, learn it vigorously and you won't struggle with positioning in CSS.

Conclusion

Thank you for reading 🔥🔥🔥, and I hope you found this article useful 🔥

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

5 Website to learn and improve your Algorithm and Data Structure skills

Favor Onuoha — Mon, 31 May 2021 10:48:39 +0000

Having extensive knowledge of Algorithms and Data Structures is a very necessary skill in our everyday programming lives.

The problem most beginners often face is the bane of finding good resources on the subject matter.

I have prepared a list of five (5) platforms that will help you improve your Algorithms and Data Structure skills.

Why Improve your Algorithmic skills?

Better chance of getting a good-paying job at a Top Tech company

Expands your codebase knowledge beyond a singular monolith style of thinking

Makes you write better code

Let's get started 🔥🔥🔥🔥

1] Hackerrank

I ❤️ Hackerrank.

Hackerrank was where I started my Algorithms and Data Structures journey over 5 years ago. Before then, I was just another moderate programmer who just wrote code anyhow he saw fit.

Hackerrank has a lot of coding challenges, tutorials, and competitions that keep you at your feet and wanting more.

The more you learn on their platform and solve challenges, the more your chances of even getting hired.

2] Topcoder

Topcoder is another great platform to learn and improve your Algorithmic skills.

Topcoder offers a lot of coding challenges and is a very competitive platform for mid to expert-level programmers. They also offer a lot of community-level tutorials for beginners seeking to improve their algorithmic skills.

3] CodeChef

CodeChef is a platform filled with a lot of active competitive programmers. In retrospect to Topcoder, CodeChef is a better alternative to start with.

CodeChef offers many great tutorials on learning Algorithms and Data Structures and is also a great place to team up with groups of like-minded programmers just like yourself.

4] Coderbyte

Coderbyte is another great platform for improving your Algorithmic skills and general programming skills.

Coderbyte offers various coding challenges and community solutions for you to learn from.

However, not a free platform, you will need to pay a subscription fee to gain access to more advanced coding challenges. However, there are free challenges available for you to begin with and they span from beginner level to expert level

5] LeetCode

LeetCode is one of the best platforms to help you enhance your skills. In retrospect to Coderbyte, LeetCode is a better alternative to start with. Just like Coderbyte, it's not entirely free and they offer a premium subscription fee to gain access to more features.

Conclusion

Thank you for reading 🔥🔥🔥, and I hope you found this article useful 🔥

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

Cheers!!!

A Simple CSS Flexbox Cheat Sheet

Favor Onuoha — Thu, 27 May 2021 16:41:48 +0000

Flexbox is amazing and they make positioning in CSS much, much better than raw positioning (Absolute, Relative, etc), but they can sometimes be very easy to forget or get mixed up with due to its structural nature.

Both Pros and Beginners struggle with keeping all the structural differences and style properties at heart so that is why I will be breaking down in sections this great cheat sheet by Joni Balogna which I found very great and amazing.

Flexbox Structure

Here we are shown the structure of flexboxes and how their origins are labeled.

We see the Main Axis (Horizontal Axis) and the Cross Axis (Vertical Axis)

Rows and Columns [Orientation]

Here we are shown the properties that activate the flexbox styles and the options we have for display orientation.

Positioning On Main Axis

Here we are shown how to position flex items on the Main Axis. Where do you want it? start, center, end?

Space Management

Here we are shown how to manage variable space of our items and how we want them displayed in a container. Based on our viewport, we can set how we want them to align, either on the same line or overflow.

Positioning On Cross Axis

Here we are shown how to position flex items on the Main Axis. We could either set our flex items to stretch across the axis or start at the end or center.

Row Management

Here we are shown the property options we have when we have multiple rows and how we can position them in a container.

Item Order

Here we are shown the order property which sets the order to lay out an item in a flex container. Items in a container are sorted by ascending order value and then by their source code order.

Item Growth

Here we are shown the flex-grow property which sets the flex-grow factor of a flex item's main size.

Alignment

Here we are shown various alignment flex properties.

Conclusion

Thank you for reading 🔥🔥🔥, and I hope you found this article useful 🔥

Here is the link to the full cheat cheat

If you have any questions or your just wanna chat, feel free to DM me

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

References

A Simple Machine Learning Cheat Sheet For Cracking The Theory Interview

Favor Onuoha — Mon, 24 May 2021 11:12:14 +0000

Hello Machine Learning Enthusiasts, I found this great Machine Learning cheat sheet that you can use for acing your ML theory interviews.

Happy Reading!!! 🔥🔥🔥

Bias vs. Variance

Imbalanced Data in Classification

PCA Dimensionality Reduction

Bayes Theorem and Classifier

Regression Analysis

Regularization in ML

Famous CNN's

Convolutional Neural Networks

Ensemble Learning in ML

How To Prepare For Behavioral Interview ? (1/4)

How To Prepare For Behavioral Interview ? (2/4)

How To Prepare For Behavioral Interview ? (3/4)

How To Prepare For Behavioral Interview ? (4/4)

Conclusion

Thank you for reading 🔥🔥🔥, and I hope you found this article useful 🔥

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

Cheers!!!

10 Important Machine Learning Terminologies you should always keep in mind

Favor Onuoha — Sun, 23 May 2021 19:57:30 +0000

It is my strong believe that the key to understanding ML is knowing its terminologies and also knowing what you are doing.

1) Epoch

A full training pass over the entire dataset such that each example has been seen once. Thus, an epoch represents N/batch size training iterations, where N is the total number of examples.

2) Example

One row of a dataset. An example contains one or more features and possibly a label. See also labeled example and unlabeled example.

3) Label

It is the result portion of an example. Each example in a dataset consists of one or more features and a label.

In a spam detection dataset, the features may include the subject, sender, and the message, while the label would probably be either "spam" or "not spam."

4) Learning Rate

A scalar used to train a model via gradient descent. During each iteration, the gradient descent algorithm multiplies the learning rate by the gradient. The resulting product is called the gradient step.

5) Model

The representation of what a machine learning system has learned from the training data.

6) Training

The process of determining the ideal parameters comprising a model.

7) Stochastic Gradient Descent (SGD)

A gradient descent algorithm in which the batch size is one. In other words, SGD relies on a single example chosen uniformly at random from a dataset to calculate an estimate of the gradient at each step.

8) Squared Loss (L2 Loss)

This function calculates the squares of the difference between a model's predicted value for a labeled example and the actual value of the label.

Due to squaring, this loss function amplifies the influence of bad predictions.

9) Loss

A measure of how far a model's predictions are from its label. Or, to phrase it more pessimistically, a measure of how bad the model is. To determine this value, a model must define a loss function.

10) Machine Learning

Some novices don't know the meaning of this term. However, they aren't to blame as to most of them simply get straight to coding and not learning the basics.

Machine Learning is a program or system that builds (trains) a predictive model from input data.

The system uses the learned model to make useful predictions from new (never-before-seen) data drawn from the same distribution as the one used to train the model.

Machine learning also refers to the field of study concerned with these programs or systems.

Conclusion

Thank you for reading 🔥🔥🔥, and I hope you found this article useful 🔥

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

Cheers!!!

How to get started with Machine Learning in 2021 [Roadmap]

Favor Onuoha — Sun, 23 May 2021 15:27:24 +0000

The process of learning Machine Learning (ML) can often be challenging and confusing. A good majority of ML enthusiast start off energetic only to give up at the start and trust me, I know and I’ve been there. I have personally gone through that daunting process not once but thrice.

I started learning Machine Learning 5 years ago. Back then, good Machine Learning resources were very limited and the best places to find quality ML resources were on sites like Udacity during the days when the term “ML” was mostly coined with the term “Data Science”. After 2 or 3 ML videos, I gave up and started learning something else. I tried again 3 years later and again, I gave up.

But what changed in 2021 ? Well for starters, I gained more experience in engineering and continued to improve upon myself and incase you were wondering, 5 years ago I was 17 years old. The second and most important reason as to how I was able to dive into Machine Learning was that I had a good roadmap and today I’ll be laying out to you my roadmap for getting started with Machine Learning.

Before I continue, I would like to state a few facts:

A solid background in Math is necessary but not required to learn ML.

A good understanding of Data Structures and Algorithms are necessary to learn ML.

A solid programming background with in-depth knowledge of programming concepts is a must have to begin with ML.

Now that's’ out of the way, lets continue.

Learn Python

Python is the leading programming language in Machine Learning development.

ML Developers love python because of its accessibility and simple uncomplicated syntax.

Though other programming languages like JavaScript are also gaining heads in the Machine Learning world, Python is by far the most used and most accessible choice and finding good resources on Python based ML codebase will not be difficult for you as a beginner.

Here is a great tutorial to begin with if you don’t already know Python.

Take Google’s Machine Learning Crash Course

Google’s Machine Learning Crash Course is by far the best crash course on ML that I have seen till date. It gets you started with basic ML principles and as well as give you a good understanding of the math involved in some of MLs core algorithms.

The crash course also teaches NumPy, Pandas and most importantly TensorFlow which is an amazing library for handling most Machine Learning related problems and packs into it all the math and algorithms needed to build a ML Model.

Optionally, after taking Google’s Machine Learning Crash Course, you can also take Coursera’s ML Course by Andrew Ng.

Learn Some Math!

Although as mentioned above that math is not required to learn Machine Learning, a good understanding of how various math calculations work in ML will give you an edge in your journey learning Machine Learning.

However, you don’t need to know all the major math topics as to only a mindful of math topics is used in Machine Learning and a few examples are :

Algebra
Linear Regression
Probability and Statistics

Probability and Statistics make up a huge chunk of the Machine Learning math in most ML libraries that are used for creating models, so those topics are where you would be focusing most of your time on.

Khan Academy has a great YouTube tutorial playlist for these particular math topics.

Learn Notebooks

Notebooks like the Jupyter Notebook are interactive multi-purpose tools that not only let you write and execute code but, at the same time, analyze intermediate results to gain insights while working on a project. Simply put, they are amazing and they speed up development.

Google Collaboratory (COLABS) is Google’s version of Jupyter’s Notebook. It creates a virtual environment in which you can write and test your code and the best part is that it is fully online and you don’t have to worry about GPU’s and Hardware specs.

Learning how to use Notebooks is a great asset to have in your journey as you learn ML.

Work with Kaggle

Kaggle is the world’s largest data science community with powerful tools and resources to help you achieve your data science goals. It was created and is maintained by Google and is a great platform for learning Machine Learning by practice.

Kaggle is home to various and popular real world scenario projects that you can work on to improve your ML skills and equip you with enough knowledge to be able to build and understand real world datasets and models on your own.

Conclusion

Learning Machine Learning has never been more easier in any other period in time except for now. There are hundreds if not thousands of learning materials out there which often just clouds your head and can make you give up prematurely but with a good roadmap, you can learn ML faster and most importantly, you can learn ML the right way!

Thank you for reading, and I hope this roadmap I have set out for you helps you navigate the cloudy path of learning Machine Learning.

If you enjoyed reading, please consider following me here on Dev.to and also on Twitter @ heyOnuoha so you don’t miss out on new content on topics like:

Web Development (JS and Python)
Improving as a developer
Machine Learning

Support Me :

👉 Follow Me on Twitter
👉 Buy me a ☕

Cheers!!!