Electronic Theses and Dissertation

Reaktivasi aset industri yang tidak beroperasi merupakan strategi esensial dalam optimalisasi portofolio aset, terutama pada sektor petrokimia yang sarat modal dan berisiko tinggi. pt pupuk iskandar muda (pt pim) dalam kerangka pengembangan kawasan ekonomi khusus arun lhokseumawe menghadapi tantangan strategis dalam memilih skema operasional paling efisien dan bernilai tambah untuk mengaktifkan kembali pabrik hydrogen peroxide (h₂o₂) eks-pt aaf. tiga skema dianalisis sewa, kerja sama operasi (kso) dan swakelola yang masing-masing memiliki implikasi berbeda terhadap struktur kontrol, potensi keuntungan dan eksposur risiko. pendekatan berbasis intuisi tidak mencukupi, diperlukan kerangka evaluasi yang mampu mengintegrasikan indikator ekonomi dan risiko secara sistematis dan berbasis data. penelitian ini bertujuan mengevaluasi kelayakan operasional melalui indikator net present value (npv), internal rate of return (irr) dan payback period (pp) serta dua dimensi risiko utama risiko operasional dan risiko pasar dengan memanfaatkan algoritma decision tree c4.5 sebagai alat klasifikasi. data sekunder dari studi kelayakan pt pim dikonversi ke format kategorikal untuk kemudian dianalisis melalui perhitungan entropy, information gain, dan gain ratio guna membentuk struktur pohon keputusan. hasil analisis menunjukkan bahwa risiko pasar merupakan atribut paling dominan dalam penentuan klasifikasi kelayakan, mengindikasikan bahwa eksposur terhadap dinamika eksternal lebih kritikal dibandingkan efisiensi internal. npv menjadi indikator ekonomi kunci yang membedakan antara keputusan “layak” dan “sangat layak”. dari ketiga skema, swakelola teridentifikasi sebagai opsi paling unggul dengan npv sebesar ±rp 38,77 miliar, irr 14,48% dan payback period 5,58 tahun serta keunggulan dalam kontrol risiko operasional. model pohon keputusan yang dibangun tidak hanya berfungsi sebagai alat klasifikasi, tetapi juga sebagai instrumen strategis untuk memahami interaksi multivariat antar indikator. temuan ini membuktikan bahwa algoritma c4.5 mampu menghasilkan kerangka pengambilan keputusan yang preskriptif, akurat, dan aplikatif bagi manajemen aset industri berbasis risiko. kata kunci: c4.5, skema operasional, pengambilan keputusan, npv, risiko pasar, reaktivasi pabrik, h₂o₂.

The The reactivation of idle industrial assets constitutes a strategic imperative in asset portfolio optimization, particularly in the capital-intensive and risk-exposed petrochemical sector. PT Pupuk Iskandar Muda (PT PIM), within the development framework of the Arun Lhokseumawe Special Economic Zone, is confronted with a critical decision in selecting the most efficient and value-generating operational scheme to recommission the decommissioned Hydrogen Peroxide (H₂O₂) plant formerly operated by PT AAF. Three alternative schemes are evaluated—leasing, joint operation (KSO), and self-operation each characterized by distinct financial structures, degrees of operational control, and risk profiles. In such a complex context, intuition-based decision-making is insufficient; a structured analytical approach is essential to integrate financial viability and risk dimensions objectively. This study aims to assess the operational feasibility of each scheme using economic indicators—Net Present Value (NPV), Internal Rate of Return (IRR), and Payback Period (PP)—and two principal risk factors: operational and market risk. The C4.5 decision tree algorithm is employed as the classification framework. Secondary data sourced from PT PIM’s internal feasibility studies were transformed into categorical variables, followed by entropy, information gain, and gain ratio computations to construct the decision tree model. The results indicate that market risk is the most dominant attribute influencing feasibility classification, underscoring the critical impact of external volatility over internal technical efficiency. NPV serves as the primary economic discriminator distinguishing between the “Feasible” and “Highly Feasible” classes. Among the evaluated options, self-operation emerged as the most favorable scheme, with a projected NPV of approximately IDR 38.77 billion, IRR of 14.48%, and a Payback Period of 5.58 years. Furthermore, it offers superior operational risk control through integrated internal management. The resulting decision tree model functions not only as a classification tool but also as a strategic instrument for elucidating multivariate interactions among key decision variables. This research validates the application of the C4.5 algorithm in delivering prescriptive, data-driven insights for investment decision-making in industrial asset management. The methodology demonstrated herein is replicable and adaptable to similar high-stakes industrial contexts. Keywords: C4.5 algorithm, decision tree, operational schemes, investment feasibility, market risk, NPV, H₂O₂ plant reactivation.