{
  "id": "uas-ai-2025",
  "subject": "Kecerdasan Artifisial",
  "exam": "UAS",
  "year": 2025,
  "variant": null,
  "sources": [
    {
      "file": "UAS/UAS - AI - 2025_1.jpg",
      "type": "image",
      "part": 1
    },
    {
      "file": "UAS/UAS - AI - 2025_2.jpg",
      "type": "image",
      "part": 2
    }
  ],
  "pageCount": 2,
  "charCount": 2770,
  "pages": [
    {
      "source": "UAS/UAS - AI - 2025_1.jpg",
      "page": 1,
      "method": "ocr",
      "text": "7 N\n|\n/\n\\ \\ »\nLi (wl MN\n— : y 8\ny Ed ~-\n~~ PROGI AN STUDI S1 ILMU KOMPLEK\nDO rn EEA Ras\nww UNIVERSITAS GADJAH MADA\nFinal F.xam\nOdd Semester 2024/2025\nCourse Name : Artificial Intelligence (CSB)\n| Day, Date : Manday. 2 December 2024\n| Time 3 1010 minutes\n| Lecturer : Dzikei Rabalian Fudholi, S.Kom., M.Comp. Se\n| Exam mode : Open one AH/F4 Sheet\n| Follow the examiner rules and good Tuck!\n8 | 1. PLO 2 & 3-CO 4-Ahle to explain Expert System Architecture (7%) 4\ni A. Explain the role of the owledge base, inference engine, and user interface in an\n| expert system! Illustrate your explanation With an example of how these components\n| neat in a specific domain (c.g, medical diagnosis or financial plannin.)\n|\n| B. Discuss the challenges in maintaining the knowledge bi c\n| 2. PLO2&3-CO 5- Able to explain Natural Language Processing system (8%)\n| A. Discuss the importance of text preprocessing in NLP tasks.\n| B. Describe the steps involved in creating a preprocessing pipeline for an emotion ion\n| recognition system, addressing challenges such as Licdling noisy data and nization in\n| morphologically rich languages.\n| : : il \\\n| C. Explain the traditional word representations of TF-IDF! Discuss why this method will\n| not perform well.\n| 3. PLO4&5-CO 6-Able to explain Pattern Recognition and Machine Learning\n| Methods (10 26)\nAn e-commerce company wants to develop a system to classify cuir into \"High-\nRisk 1 ow-Risk™ gro for fraudulent transactions based on numerical (-nives. The\ndataset Includes transact: amounts, frequency of purchases, account age, 1 nd rates,\nAnd tier behaviour metrics. The goal is to accurately classify customers while\nminimizing false positives (lagging legitimate customers as high-risk) and false negatives\n(missing fraudulent customers).\nTask:\nWrite an essay proposal prefer ly in points, addressing the following:\n| Telah direview pada tanggal: cc... Tanda tangan Reviewer:\nNama ReVIEWEr: ...oooocosesinin,\nbh _ ab\nrr - |\nDu” |\nAa binaan aa\nPEN ~\ni RN “ rr AM\nAA A\n5 KS ,\nAd iis. ag\nAMI\nVi { eS kK ON\nkN . ."
    },
    {
      "source": "UAS/UAS - AI - 2025_2.jpg",
      "page": 1,
      "method": "ocr",
      "text": "Dn / \\\nSal IK V\nmm —\"\n—\n1. Data Exploration and Preprocessing: .\nDiscuss how to handle missing values, outliers, and skewed distributions in 3\nnumerical features, >\n© Propose techniques for feature selection or tity reduction to\nenhance model interpretability and pertormance\n2. Model Selection and Training:\n© Compare classification models for this task.\n© Explain how you would tune hyper parameter to optimize classification\nperformance\n3. Evaluation and Chew balance: 4\n© Identify appropriate evaluation metrics and justify your choices.\n© Discuss methods to handle imbalanced classes.\n| >>> Best of Luck to you Ay <<<\n|\n1\n|\nTER —\nale To Tanda ago Rena\na SW pada tanggal: osssssssnss..,.\nd [an ey\nowe,\nA A 2 Pin agg 4 4"
    }
  ],
  "text": "7 N\n|\n/\n\\ \\ »\nLi (wl MN\n— : y 8\ny Ed ~-\n~~ PROGI AN STUDI S1 ILMU KOMPLEK\nDO rn EEA Ras\nww UNIVERSITAS GADJAH MADA\nFinal F.xam\nOdd Semester 2024/2025\nCourse Name : Artificial Intelligence (CSB)\n| Day, Date : Manday. 2 December 2024\n| Time 3 1010 minutes\n| Lecturer : Dzikei Rabalian Fudholi, S.Kom., M.Comp. Se\n| Exam mode : Open one AH/F4 Sheet\n| Follow the examiner rules and good Tuck!\n8 | 1. PLO 2 & 3-CO 4-Ahle to explain Expert System Architecture (7%) 4\ni A. Explain the role of the owledge base, inference engine, and user interface in an\n| expert system! Illustrate your explanation With an example of how these components\n| neat in a specific domain (c.g, medical diagnosis or financial plannin.)\n|\n| B. Discuss the challenges in maintaining the knowledge bi c\n| 2. PLO2&3-CO 5- Able to explain Natural Language Processing system (8%)\n| A. Discuss the importance of text preprocessing in NLP tasks.\n| B. Describe the steps involved in creating a preprocessing pipeline for an emotion ion\n| recognition system, addressing challenges such as Licdling noisy data and nization in\n| morphologically rich languages.\n| : : il \\\n| C. Explain the traditional word representations of TF-IDF! Discuss why this method will\n| not perform well.\n| 3. PLO4&5-CO 6-Able to explain Pattern Recognition and Machine Learning\n| Methods (10 26)\nAn e-commerce company wants to develop a system to classify cuir into \"High-\nRisk 1 ow-Risk™ gro for fraudulent transactions based on numerical (-nives. The\ndataset Includes transact: amounts, frequency of purchases, account age, 1 nd rates,\nAnd tier behaviour metrics. The goal is to accurately classify customers while\nminimizing false positives (lagging legitimate customers as high-risk) and false negatives\n(missing fraudulent customers).\nTask:\nWrite an essay proposal prefer ly in points, addressing the following:\n| Telah direview pada tanggal: cc... Tanda tangan Reviewer:\nNama ReVIEWEr: ...oooocosesinin,\nbh _ ab\nrr - |\nDu” |\nAa binaan aa\nPEN ~\ni RN “ rr AM\nAA A\n5 KS ,\nAd iis. ag\nAMI\nVi { eS kK ON\nkN . .\n\nDn / \\\nSal IK V\nmm —\"\n—\n1. Data Exploration and Preprocessing: .\nDiscuss how to handle missing values, outliers, and skewed distributions in 3\nnumerical features, >\n© Propose techniques for feature selection or tity reduction to\nenhance model interpretability and pertormance\n2. Model Selection and Training:\n© Compare classification models for this task.\n© Explain how you would tune hyper parameter to optimize classification\nperformance\n3. Evaluation and Chew balance: 4\n© Identify appropriate evaluation metrics and justify your choices.\n© Discuss methods to handle imbalanced classes.\n| >>> Best of Luck to you Ay <<<\n|\n1\n|\nTER —\nale To Tanda ago Rena\na SW pada tanggal: osssssssnss..,.\nd [an ey\nowe,\nA A 2 Pin agg 4 4"
}