dineshb's picture
Deploy DataPilot AI production Docker Space
9c1c0ef verified
Raw
History Blame Contribute Delete
2.3 kB
<!doctype html>
<html lang="en"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width">
<title>DataPilot AI report</title>
<style>
body{font-family:Inter,system-ui,sans-serif;max-width:1000px;margin:40px auto;padding:0 24px;color:#172033}
h1{color:#5537d8} .hero{background:#f4f1ff;border:1px solid #d9d0ff;padding:24px;border-radius:18px}
.grid{display:grid;grid-template-columns:repeat(auto-fit,minmax(220px,1fr));gap:16px;margin:20px 0}
.card{border:1px solid #e2e6ef;border-radius:14px;padding:18px} table{border-collapse:collapse;width:100%}
td,th{border-bottom:1px solid #e2e6ef;padding:10px;text-align:left} small{color:#667085}
</style></head><body>
<div class="hero"><h1>DataPilot AI Analysis Report</h1>
<p>iris · Run run_f13555d6077f</p></div>
<div class="grid">
<div class="card"><small>Selected model</small><h2>Logistic Regression</h2></div>
<div class="card"><small>Test balanced_accuracy</small><h2>0.933</h2></div>
<div class="card"><small>Rows analyzed</small><h2>150</h2></div>
</div>
<h2>Executive findings</h2><ul><li>The analysis used 150 rows and 5 columns for a classification task targeting &#x27;target&#x27;.</li><li>Logistic Regression ranked first with training CV balanced_accuracy 0.958; its one-time test score was 0.933.</li><li>1 data-quality observations were recorded; 0 are critical.</li><li>The strongest predictive signals were petal_width_(cm), petal_length_(cm), sepal_length_(cm) according to permutation importance.</li></ul>
<h2>Evaluation</h2><table><tr><th>Metric</th><th>Value</th></tr><tr><td>accuracy</td><td>0.9333</td></tr><tr><td>balanced_accuracy</td><td>0.9333</td></tr><tr><td>f1_weighted</td><td>0.9333</td></tr></table>
<h2>Data quality</h2><table><tr><th>Severity</th><th>Code</th><th>Observation</th></tr><tr><td>warning</td><td>DUPLICATE_ROWS</td><td>1 exact duplicate rows can bias validation.</td></tr></table>
<h2>Recommendations</h2><ol><li>Validate performance on fresh, out-of-time data before production deployment.</li><li>Review suspected leakage and identifier columns with a domain owner.</li><li>Monitor input drift and the primary metric after deployment.</li></ol>
<p><small>Generated from computed evidence. Predictive findings do not establish causality.</small></p>
</body></html>