Skip to content

Commit 1899e3c

Browse files
DataForSolutionDataForSolution
authored
docs: improve curated project cards
1 parent cd19ab0 commit 1899e3c

1 file changed

Lines changed: 44 additions & 27 deletions

File tree

‎projects.md‎

Lines changed: 44 additions & 27 deletions
Original file line numberDiff line numberDiff line change
@@ -23,66 +23,66 @@ description: "Selected healthcare AI, responsible AI, machine learning, NLP, and
2323
<h2>Healthcare &amp; Medical Imaging</h2>
2424
<div class="project-grid">
2525
<article class="project-card">
26-
<h3>PETQuant Reliability</h3>
26+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/petquant" target="_blank" rel="noopener">PETQuant Reliability</a></h3>
2727
<p>PET quantitative-ML reconstruction focused on small-sample limits, data-quality contracts, and defensible validation claims.</p>
28-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/petquant" target="_blank" rel="noopener">View project</a>
28+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>Python</li><li>PET safety</li><li>Validation</li></ul>
2929
</article>
3030
<article class="project-card">
31-
<h3>Chest CT Classification</h3>
31+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/chest-ct-classification" target="_blank" rel="noopener">Chest CT Classification</a></h3>
3232
<p>Medical-imaging transfer-learning reconstruction emphasizing overfitting analysis, split integrity, and deterministic evaluation.</p>
33-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/chest-ct-classification" target="_blank" rel="noopener">View project</a>
33+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>TensorFlow</li><li>ResNet50</li><li>Medical imaging</li></ul>
3434
</article>
3535
<article class="project-card">
36-
<h3>Pima Diabetes ML Evaluation</h3>
36+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/pima-diabetes-ml" target="_blank" rel="noopener">Pima Diabetes ML Evaluation</a></h3>
3737
<p>Leakage-safe SVM/MLP retrospective with clinically relevant missing-measurement handling and sensitivity/specificity reporting.</p>
38-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/pima-diabetes-ml" target="_blank" rel="noopener">View project</a>
38+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>scikit-learn</li><li>Clinical metrics</li><li>Leakage control</li></ul>
3939
</article>
4040
</div>
4141

4242
<h2>Responsible &amp; Explainable AI</h2>
4343
<div class="project-grid">
4444
<article class="project-card">
45-
<h3>SHAP + LIME Explainability</h3>
45+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/shap-lime-explainability" target="_blank" rel="noopener">SHAP + LIME Explainability</a></h3>
4646
<p>Class-aligned explanation workflow with corrected label ordering, scaled modeling, and defensible attribution aggregation.</p>
47-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/shap-lime-explainability" target="_blank" rel="noopener">View project</a>
47+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>SHAP</li><li>LIME</li><li>Explainability</li></ul>
4848
</article>
4949
<article class="project-card">
50-
<h3>Fairness Evaluation</h3>
50+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/fairness-evaluation" target="_blank" rel="noopener">Fairness Evaluation</a></h3>
5151
<p>Explicit group-fairness metrics and threshold analysis with fixed reference labels and documented AIF360 provenance.</p>
52-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/fairness-evaluation" target="_blank" rel="noopener">View project</a>
52+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>AIF360</li><li>Group fairness</li><li>Thresholds</li></ul>
5353
</article>
5454
<article class="project-card">
55-
<h3>Adversarial Robustness</h3>
55+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/adversarial-robustness" target="_blank" rel="noopener">Adversarial Robustness</a></h3>
5656
<p>Bounded PGD and defense-evaluation reconstruction with preprocessing, threat-model, and denominator corrections.</p>
57-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/adversarial-robustness" target="_blank" rel="noopener">View project</a>
57+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>PyTorch</li><li>PGD</li><li>Threat models</li></ul>
5858
</article>
5959
</div>
6060

6161
<h2>Applied Machine Learning &amp; NLP</h2>
6262
<div class="project-grid">
6363
<article class="project-card">
64-
<h3>CIFAR-10 Generated-Image Analysis</h3>
65-
<p>Classifier-response diagnostics for generated images with a clear separation between classifier confidence and perceptual realism.</p>
66-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/cifar10-generated-analysis" target="_blank" rel="noopener">View project</a>
67-
</article>
68-
<article class="project-card">
69-
<h3>Automobile Insurance Fraud Classification</h3>
64+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/insurance-fraud-classification" target="_blank" rel="noopener">Automobile Insurance Fraud Classification</a></h3>
7065
<p>Temporal, imbalance-aware evaluation showing why high overall accuracy can coexist with zero minority-class recall.</p>
71-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/insurance-fraud-classification" target="_blank" rel="noopener">View project</a>
66+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>PR-AUC</li><li>Temporal holdout</li><li>Imbalanced ML</li></ul>
7267
</article>
7368
<article class="project-card">
74-
<h3>Restaurant Review Sentiment Evaluation</h3>
69+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/restaurant-sentiment-evaluation" target="_blank" rel="noopener">Restaurant Review Sentiment Evaluation</a></h3>
7570
<p>Leakage-safe NLP model comparison using fold-local TF-IDF pipelines and a clearly licensed UCI Yelp data source.</p>
76-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/restaurant-sentiment-evaluation" target="_blank" rel="noopener">View project</a>
71+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>TF-IDF</li><li>NLP</li><li>Cross-validation</li></ul>
72+
</article>
73+
<article class="project-card">
74+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/cifar10-generated-analysis" target="_blank" rel="noopener">CIFAR-10 Generated-Image Analysis</a></h3>
75+
<p>Classifier-response diagnostics for generated images with a clear separation between classifier confidence and perceptual realism.</p>
76+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>NumPy</li><li>Entropy</li><li>Generative AI</li></ul>
7777
</article>
7878
</div>
7979

8080
<h2>Data Engineering</h2>
8181
<div class="project-grid">
8282
<article class="project-card">
83-
<h3>AWS Glue + EMR PySpark Ingestion Validation</h3>
83+
<h3><a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/spark-glue-ingestion" target="_blank" rel="noopener">AWS Glue + EMR PySpark Ingestion Validation</a></h3>
8484
<p>Catalog-resolution and ingestion-quality engineering based on a historical embedded-header defect discovered in an EMR/Glue run.</p>
85-
<a href="https://github.com/DataForSolution/Data-Solution/tree/Master/portfolio/spark-glue-ingestion" target="_blank" rel="noopener">View project</a>
85+
<ul class="project-tags" aria-label="Core technologies and capabilities"><li>PySpark</li><li>AWS Glue</li><li>Data quality</li></ul>
8686
</article>
8787
</div>
8888
</section>
@@ -122,16 +122,33 @@ description: "Selected healthcare AI, responsible AI, machine learning, NLP, and
122122
.project-card h3 {
123123
margin-top: 0;
124124
}
125+
.project-card h3 a {
126+
color: #123b63;
127+
text-decoration-thickness: 2px;
128+
text-underline-offset: 0.18em;
129+
}
125130
.project-card p {
126131
color: #1f2937;
127132
line-height: 1.6;
128133
}
129-
.project-card a {
130-
display: inline-block;
131-
margin-top: 0.5rem;
134+
.project-tags {
135+
display: flex;
136+
flex-wrap: wrap;
137+
gap: 0.45rem;
138+
margin: 1rem 0 0;
139+
padding: 0;
140+
list-style: none;
141+
}
142+
.project-tags li {
143+
border: 1px solid #b9c8d8;
144+
border-radius: 999px;
145+
padding: 0.25rem 0.6rem;
146+
background: #f3f7fb;
147+
color: #1f2937;
148+
font-size: 0.85rem;
132149
font-weight: 600;
133150
}
134151
.project-note {
135152
border-top: 1px solid #ddd;
136153
}
137-
</style>
154+
</style>

0 commit comments

Comments
 (0)