<![CDATA[This study aims to determine the effectiveness of the deep learning approach on students' problem-solving skills at SMK Negeri 1 Purbolinggo. The deep learning approach was chosen because it emphasizes students' activities in understanding the material more meaningfully, integrating knowledge with real experiences, and honing students' problem-solving skills. This study used a quasi-experimental method with a pretest-posttest control group design. The research subjects consisted of two classes, namely the experimental class using the deep learning approach based on the PBL learning model, and the control class using the scientific approach based on the PBL model. Data analysis was carried out using the Independent Sample t-test and the N-Gain Score test with the help of the SPSS 25 program. The results showed that there was a significant difference between students' problem-solving skills in the experimental class and the control class. This was indicated by the Sig. (2-tailed) value of 0.002 <0.05, which means that H₀ was rejected and H₁ was accepted. The average N-Gain Score of the experimental class was 67.45, categorized as quite effective, while the control class was 60.27, categorized as quite effective. Thus, the deep learning approach through the PBL learning model has proven effective in improving students' problem-solving skills because it encourages students to understand the material more deeply, relate knowledge to real situations, and develop the ability to become accustomed to solving problems.]]>
