{
  "source_file": "0930",
  "source_platform": "Hashnode",
  "author": "Tharun",
  "title_listed": "I Thought Machine Learning Was About Code. I Was Completely Wrong.",
  "keywords": "",
  "domain": "机器学习",
  "content_type": "深度技术解析",
  "is_paid_flag": "否",
  "url": "https://tharunpandya.hashnode.dev/why-i-learn-machine-learning-from-scratch",
  "publish_date": "2026-07-03",
  "corpus_stage": "",
  "notes": "",
  "id": "3b1a545e52b4",
  "collected_at": "2026-10-02T14:10:46",
  "title": "Why I Learn Machine Learning from Scratch",
  "body_markdown": "I Thought Machine Learning Was About Code. I Was Completely Wrong.\nUpdated\nJuly 8, 2026\n•\n4\nmin read\n•\nView as Markdown\nT\nTharun\n\"Why am I learning all this mathematics?\"\nThat question haunted me every time I sat down to learn Machine Learning.\nI wasn't intimidated by Python. I enjoyed writing code, experimenting with libraries, and building small projects. But every tutorial seemed to begin with something I wasn't prepared for—linear algebra, calculus, probability, optimization, gradients, vectors, and matrices.\nI kept wondering:\n\"Do I really need all of this just to build AI?\"\nIf you've ever started learning Machine Learning, you've probably had the same thought.\nYou open a tutorial expecting to teach a computer how to think.\nInstead, you're greeted by pages of mathematical equations that seem completely disconnected from the code you're trying to write.\nSomewhere between the mathematics and the implementation, the excitement begins to fade.\nI know that feeling because I lived it.\nLike many beginners, I believed that learning Machine Learning meant learning frameworks. I spent countless hours working with Python, NumPy, Pandas, scikit-learn, and TensorFlow. My code executed successfully. My models trained. The predictions looked reasonable.\nFrom the outside, everything seemed to be working.\nBut there was one problem.\nIf someone had stopped me and asked,\n\"How does a machine actually learn?\"\nI wouldn't have known how to answer.\nI could build models.\nI could follow tutorials.\nI could reproduce projects.\nBut I couldn't explain\nwhy\nany of it worked.\nThat realization was frustrating.\nEvery course taught me\nwhat\nto write.\nVery few taught me\nwhy\nI was writing it.\nWithout realizing it, I had fallen into a common trap.\nI wasn't learning Machine Learning.\nI was learning how to use Machine Learning.\nThere's a huge difference.\nEventually, I stopped learning altogether.\nNot because Machine Learning was too difficult.\nBut because I felt like I was building something I didn't truly understand.\nMonths later, I decided to start over.\nThis time, I made myself one promise.\nI wouldn't memorize another algorithm until I understood\nwhy\nit existed.\nInstead of asking,\n\"How do I implement this algorithm?\"\nI started asking a completely different question:\n\"If I were a computer, how would I learn from experience?\"\nThat single question changed everything.\nRather than looking at Machine Learning as lines of Python code, I started comparing it with something I experience every day—the human brain.\nSuddenly, concepts that once felt overwhelming became surprisingly intuitive.\nOur experiences became\ntraining data\n.\nOur memories became\nstored knowledge\n.\nOur mistakes became\nfeedback\n.\nEvery decision we make is influenced by patterns we've learned throughout our lives, just as Machine Learning models improve by learning patterns from data.\nFor the first time, Machine Learning no longer felt like magic.\nIt felt like a mathematical explanation of how learning itself works.\nThen came an even bigger realization.\nFor years, I believed mathematics was just another academic subject—a collection of formulas I'd memorize for an exam and forget a few weeks later.\nI couldn't have been more wrong.\nMachine Learning showed me that mathematics isn't just something we study.\nIt's the language we use to describe intelligence.\nEvery recommendation on YouTube.\nEvery face recognized by your phone.\nEvery movie suggested by Netflix.\nEvery translated sentence.\nEvery AI-generated image.\nEvery prediction made by a Machine Learning model.\nBehind each of these is mathematics quietly describing patterns hidden inside data.\nThe formulas I once tried to avoid suddenly had purpose.\nI stopped memorizing equations.\nI started understanding ideas.\nAnd that changed everything.\nIt transformed the way I approach Machine Learning, Software Engineering, and learning itself.\nIt also inspired me to build Machine Learning algorithms from scratch.\nNot because libraries like TensorFlow, PyTorch, or scikit-learn aren't incredible—they absolutely are.\nBut because rebuilding the fundamentals forces you to understand why every neuron, every weight, every activation function, every loss function, and every optimization step exists.\nWhen you understand the foundations, frameworks stop feeling like magic.\nThey become tools.\nThat philosophy became the foundation of this blog.\nThis isn't a place where I'll simply show you how to use another library or copy code from a tutorial.\nInstead, we'll explore the intuition behind Machine Learning.\nWe'll build algorithms from scratch.\nWe'll connect mathematics with real-world ideas.\nWe'll break complex concepts into simple, understandable pieces.\nAnd most importantly, we'll always ask\n\"Why?\"\nbefore\n\"How?\"\nWhether you're taking your first steps into AI or you're an experienced engineer who enjoys understanding systems from first principles, I hope you'll find something valuable here.\nBecause learning Machine Learning isn't just about teaching computers to learn.\nIt's about learning how to think differently yourself.\nWelcome to the journey.\n#\nmachine-learning\n#\nartificial-intelligence\n#\npython\n#\nsoftware-development\n#\ncareer\n#\nself-improvement\n#\nmachine-learning-from-scratch\n#\ndeep-learning\n36\nviews",
  "metadata": {
    "title": "Why I Learn Machine Learning from Scratch",
    "description": "Discover how learning Machine Learning from scratch changed my perspective on mathematics, AI, and the way I approach learning.",
    "keywords": null,
    "author": null,
    "og:title": "Why I Learn Machine Learning from Scratch",
    "og:description": "Discover how learning Machine Learning from scratch changed my perspective on mathematics, AI, and the way I approach learning.",
    "og:url": "https://tharunpandya.hashnode.dev/why-i-learn-machine-learning-from-scratch",
    "og:site_name": "Machine Learing From scratch",
    "og:image": "https://cdn.hashnode.com/uploads/og-images/6a3f9a25a837e83b51cc8712/a35203b8-70d5-4e29-8235-1f4e64fdbae2.png",
    "og:image:width": "1200",
    "og:image:height": "630",
    "og:type": "article",
    "twitter:card": "summary_large_image",
    "twitter:title": "Why I Learn Machine Learning from Scratch",
    "twitter:description": "Discover how learning Machine Learning from scratch changed my perspective on mathematics, AI, and the way I approach learning.",
    "twitter:image": "https://cdn.hashnode.com/uploads/og-images/6a3f9a25a837e83b51cc8712/a35203b8-70d5-4e29-8235-1f4e64fdbae2.png",
    "article:published_time": "2026-07-03T07:51:24.650Z",
    "article:modified_time": "2026-07-08T12:37:49.431Z",
    "article:author": "https://tharunpandya.hashnode.dev/@tharun-pandya"
  },
  "paywall": {
    "listed_paid": false,
    "detected": false,
    "marker": "",
    "public_only": false,
    "policy": "付费/会员内容只保留页面公开可见部分，不绕过付费墙"
  },
  "crawl": {
    "success": true,
    "status_code": 200,
    "error": ""
  },
  "code_blocks": [],
  "images": [
    {
      "status": "ok",
      "url": "https://cdn.hashnode.com/uploads/covers/6a3f9a25a837e83b51cc8712/539255f3-b0ad-4d4a-a859-fd73af335330.png",
      "bytes": 1698571,
      "reused": true,
      "local_path": "attachments/3b1a545e52b4/images/002.png",
      "src": "https://cdn.hashnode.com/uploads/covers/6a3f9a25a837e83b51cc8712/539255f3-b0ad-4d4a-a859-fd73af335330.png",
      "alt": "I Thought Machine Learning Was About Code. I Was Completely Wrong.",
      "downloaded": true
    }
  ],
  "audio": [],
  "videos": [],
  "links": [
    {
      "href": "https://tharunpandya.hashnode.dev/why-i-learn-machine-learning-from-scratch.md",
      "label": "View as Markdown",
      "type": "a"
    },
    {
      "href": "https://hashnode.com/@tharun-pandya",
      "label": "T",
      "type": "a"
    },
    {
      "href": "https://linkedin.com/in/tharun-pandya-kodi",
      "label": "",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/machine-learning",
      "label": "# machine-learning",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/artificial-intelligence",
      "label": "# artificial-intelligence",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/python",
      "label": "# python",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/software-development",
      "label": "# software-development",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/career",
      "label": "# career",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/self-improvement",
      "label": "# self-improvement",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/machine-learning-from-scratch",
      "label": "# machine-learning-from-scratch",
      "type": "a"
    },
    {
      "href": "https://tharunpandya.hashnode.dev/tag/deep-learning",
      "label": "# deep-learning",
      "type": "a"
    }
  ],
  "retry_at": "2026-10-03T08:05:20",
  "repair_version": 2,
  "repaired_at": "2026-10-14T10:50:32Z",
  "source_url": "https://tharunpandya.hashnode.dev/why-i-learn-machine-learning-from-scratch",
  "papers": [],
  "actual_video_files": [],
  "body_chars": 5249,
  "expected_body_images": 1,
  "discovery_status": "ok",
  "delivery_status": "ok",
  "missing_reason": "",
  "raw_html": "attachments/3b1a545e52b4/pages/source.html",
  "local_html": "attachments/3b1a545e52b4/pages/article.html",
  "body_base_path": "attachments/3b1a545e52b4/pages/",
  "body_markdown_path": "attachments/3b1a545e52b4/pages/article.md"
}